99-ModelGrader_IndivRF_ModelPicking_Figures_NWMAdded.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "fe7ec2cb",
   "metadata": {},
   "outputs": [],
   "source": [
    "# ESSENTIALS\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import glob, os\n",
    "from datetime import datetime\n",
    "\n",
    "# CLUSTERING AND RANDOM FOREST\n",
    "import skfuzzy as fuzz\n",
    "from sklearn import tree\n",
    "from sklearn import model_selection\n",
    "from sklearn.datasets import make_blobs\n",
    "from sklearn.base import BaseEstimator, RegressorMixin\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.inspection import permutation_importance\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.svm import *\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.tree import export_graphviz\n",
    "from sklearn.metrics import mean_squared_error as mse\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.metrics import ConfusionMatrixDisplay\n",
    "\n",
    "# DATA LIBRARIES\n",
    "# import rasterio\n",
    "# import contextily as cx\n",
    "import geopandas as gpd\n",
    "import pandas as pd\n",
    "import json\n",
    "import pickle\n",
    "import xarray\n",
    "from xarray.core.dataarray import DataArray\n",
    "\n",
    "\n",
    "# PREFERENCES\n",
    "pd.set_option('display.max_columns', 500)\n",
    "plt.rcParams['figure.dpi'] = 400\n",
    "plt.rcParams['figure.figsize'] = (10, 10)\n",
    "plt.rcParams.update({'font.size': 14})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "df641318",
   "metadata": {},
   "outputs": [],
   "source": [
    "class ModelSelector():\n",
    "\n",
    "    def __init__(self, c_kwargs={}, rf_kwargs={}):\n",
    "       self.c_kwargs=c_kwargs        # CLUSTERING HYPERPARAMETERS\n",
    "       self.rf_kwargs=rf_kwargs      # RANDOM FOREST HYPERPARAMETERS\n",
    "       self.m = 2                    # EXPONENTIATION COEFFICIENT FOR CLUSTERING. TODO: MAKE ADJUSTABLE\n",
    "\n",
    "    def fuzzyCluster(self, data):\n",
    "        # Wraps Fuzzy Cluster function, only outputting percent belongs and formal cluster.\n",
    "\n",
    "        # CHECK THAT REQUIRED FIELDS ARE IN KWARGS, IF NOT ADD\n",
    "        if \"error\" not in self.c_kwargs:\n",
    "            self.c_kwargs['error']=0.005\n",
    "\n",
    "        if \"maxiter\" not in self.c_kwargs:\n",
    "            self.c_kwargs['maxiter']=1000\n",
    "\n",
    "        cntr, u, u0, d, jm, p, fpc = fuzz.cluster.cmeans(data.T, self.n_centers, self.m, **self.c_kwargs)\n",
    "        label = np.argmax(u, axis=0)\n",
    "        return cntr, u, fpc, label\n",
    "\n",
    "    def howManyClusters(self, X, mintest=2,maxtest=15):\n",
    "        # Determines how many clusters should be used using the Fuzzy Partitions Coefficient (FPC)\n",
    "        # https://scikit-fuzzy.github.io/scikit-fuzzy/auto_examples/plot_cmeans.html#example-plot-cmeans-py\n",
    "        # TODO: FIGURE OUT IF THIS METHOD IS APPROPRIATE OR NOT\n",
    "        return 3\n",
    "        fpcs = []\n",
    "        listtests = np.arange(mintest,maxtest)\n",
    "        for ncenters in listtests:\n",
    "            self.n_centers = ncenters\n",
    "            _, _, fpc, _ = self.fuzzyCluster(X)\n",
    "            fpcs.append(fpc)\n",
    "        return listtests[np.argmax(fpcs)]\n",
    "\n",
    "    def train_rf(self, X_train, y_train, rf_controls={}):\n",
    "        # ADAPTED FROM https://stackoverflow.com/questions/28489667/combining-random-forest-models-in-scikit-learn\n",
    "\n",
    "        # RF CONTROLS PASSED DIRECTLY FROM PARAMETER, DEFAULT IS EMPTY\n",
    "        rf = RandomForestRegressor(**rf_controls) \n",
    "\n",
    "        # RF FITTING \n",
    "        rf.fit(X_train, y_train)\n",
    "\n",
    "        return rf\n",
    "\n",
    "    def fit(self, attributes, model_perf):\n",
    "\n",
    "        # CREATE RANDOM FOREST AND TRAIN\n",
    "        self.rf = self.train_rf(attributes, model_perf, rf_controls=self.rf_kwargs)\n",
    "        # print(r2_score(self.rf.predict(attributes), model_perf))\n",
    "\n",
    "        return self\n",
    "\n",
    "    def predict(self, attributes):\n",
    "\n",
    "        # CHECK WHETHER MODEL HAS BEEN TRAINED\n",
    "        if self.rf is None:\n",
    "            raise(Exception(\"ModelSelector isn't trained!\"))\n",
    "\n",
    "        # GET RANDOM FOREST PREDICTION\n",
    "        pred = self.rf.predict(attributes)\n",
    "\n",
    "\n",
    "        return pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "70e7d12a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# ADAPTED FROM https://github.com/neuralhydrology/neuralhydrology/blob/1ff36ea8c8eff99ad25fa0f56f0119acbc9e6799/neuralhydrology/evaluation/metrics.py\n",
    "def nse(obs: DataArray, sim: DataArray) -> float:\n",
    "    denominator = ((obs - obs.mean())**2).sum()\n",
    "    numerator = ((sim - obs)**2).sum()\n",
    "    value = 1 - numerator / denominator\n",
    "    return float(value)\n",
    "\n",
    "def getNNSEfromNWM():\n",
    "    \n",
    "    # DEFINE DIR WITH ALL STREAMFLOW DATA \n",
    "    q_dir = \"/home/ottersloth/data/camels_hourly/usgs_streamflow\"\n",
    "    \n",
    "    # TEST PERIOD\n",
    "    test_start=datetime.strptime(\"2002-09-30 23:00:00\", '%Y-%m-%d %H:%M:%S')\n",
    "    # ORIGINALLY UNTIL 11 PM, CHANGED BECAUSE COMPARISON FUNCTION IS INCLUSIVE LATER ON\n",
    "    test_end=datetime.strptime(\"2007-09-30 22:00:00\", '%Y-%m-%d %H:%M:%S') \n",
    "    \n",
    "    # DEFINE AND READ IN NWM TIMESERIES\n",
    "    nwmdir = \"/home/ottersloth/cfe_calibration/NWM_streamflow_results.csv\"\n",
    "    nwm = pd.read_csv(nwmdir)\n",
    "    \n",
    "    # GET BASINS IN DATASET, REMOVE FIRST COLUMN NAME BC IT'S D\n",
    "    basins = nwm.columns.to_list()[1:]\n",
    "    \n",
    "    # CONTAINER FOR OUTPUT\n",
    "    nnsedf = list()\n",
    "    \n",
    "    # LOOP THROUGH BASINS IN DATASET\n",
    "    for currbasin in basins:\n",
    "        try:\n",
    "        \n",
    "            basinid = currbasin\n",
    "\n",
    "            # GET MODEL PREDICTION IN CURRENT BASINS\n",
    "            pred = nwm[currbasin].to_numpy()\n",
    "\n",
    "            # GET Q FROM USGS FOR TEST PERIOD\n",
    "            q_read = pd.read_csv(f\"{q_dir}/{basinid}-usgs-hourly.csv\")\n",
    "            q_read[\"datetime\"] = pd.to_datetime(q_read['date'], format='%Y-%m-%d %H:%M:%S') # CONVERT TO DATETIME\n",
    "            q_match = q_read[q_read.datetime.between(test_start, test_end)]\n",
    "            ts_q = q_match['QObs_CAMELS(mm/h)'].to_numpy()\n",
    "\n",
    "            # CALCULATE NSE \n",
    "            nse_calc = nse(ts_q, pred)\n",
    "\n",
    "            # CALCULATE NNSE\n",
    "            nnse = 1 / (2 - nse_calc)\n",
    "\n",
    "            # APPEND TO CONTAINER\n",
    "            # df_add = {currbasin: nnse}\n",
    "            df_add = list(np.array([currbasin, nnse]))\n",
    "            \n",
    "            nnsedf.append(df_add)\n",
    "            print(df_add)\n",
    "            \n",
    "        \n",
    "        except: \n",
    "            continue\n",
    "            \n",
    "    nnsedf = pd.DataFrame(np.array(nnsedf), columns=[\"basin_id\", \"nnse\"])\n",
    "    nnsedf.to_csv(r\"/home/ottersloth/ensemblennse/data/data/hourly_performances/nwm.csv\")\n",
    "    \n",
    "    \n",
    "    return nnsedf\n",
    "\n",
    "# nwm_nnse = getNNSEfromNWM()\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "0dfb444b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def getFeatureImportance(selector, testidx, input, output, xx, reps=10):\n",
    "    # USING FULL PERMUTATION IMPORTANCE, AS OUTLINED IN \n",
    "    # https://scikit-learn.org/stable/auto_examples/ensemble/plot_forest_importances.html\n",
    "    result = permutation_importance(selector.rf, input.iloc[testidx, :], output.iloc[testidx], n_repeats=reps, random_state=42, n_jobs=2)\n",
    "    means = pd.Series(result.importances_mean, index=list(input)) \n",
    "    std = pd.Series(result.importances_std, index=list(input))\n",
    "    df = pd.concat([means, std], axis=1)\n",
    "    df.columns = [str(xx) + \"_mean\", str(xx) + \"_std\"]\n",
    "    return df\n",
    "\n",
    "\n",
    "def runFitMetric_getMSE(fitmet, rf_kwargs={}):\n",
    "    # FILEPATH TO SHAPEFILE CONTAINING CAMELS DATASET\n",
    "    # camelsdir = r\"C:\\Users\\franc\\OneDrive - University Of Houston\\000_SIResearch\\data\\HCDN_nhru_final\\HCDN_nhru_final_671.shp\"\n",
    "    camelsdir = r\"/home/ottersloth/ensemblennse/data/data/HCDN_nhru_final/HCDN_nhru_final_671.shp\"\n",
    "    \n",
    "    # DIRECTORY TO FOLDER CONTAINING CAMELS ATTRIBUTE TEXTFILES\n",
    "    # PRIOR TO THIS STEP MAKE SURE THE README IN THE FILE SYSTEM HAS BEEN REMOVED (or the file extension has been changed)\n",
    "    attdir = r\"/home/ottersloth/ensemblennse/data/data/camels_attributes_v2.0/camels_attributes_v2.0/\"\n",
    "\n",
    "    # READ CAMELS DATASET\n",
    "    camels = gpd.read_file(camelsdir)\n",
    "\n",
    "    # COPY TO KEEP ORIGINAL IN MEMORY\n",
    "    camels_df = camels \n",
    "\n",
    "    # LOOP THROUGH AND JOIN\n",
    "    filelist = glob.glob(attdir + \"*.txt\")\n",
    "    for i in filelist:\n",
    "        currdf = pd.read_csv(i, sep=\";\")\n",
    "        camels_df = camels_df.merge(currdf, how='left', left_on=\"hru_id\", right_on=\"gauge_id\")\n",
    "\n",
    "    # DEFINE WHAT WE WANT TO RUN ON\n",
    "    # perf_dir = r\"C:\\Users\\franc\\OneDrive - University Of Houston\\000_SIResearch\\data\\hourly_performances\\\\\"\n",
    "    perf_dir = r\"/home/ottersloth/ensemblennse/data/data/hourly_performances/\"\n",
    "    \n",
    "    perf_metrics = [fitmet]\n",
    "\n",
    "    # READ ALL CSV FILES IN DIRECTORY\n",
    "    os.chdir(perf_dir)\n",
    "    modelfiles = glob.glob(\"*.csv\")\n",
    "\n",
    "    # GET FIRST CSV FILE TO DEFINE DATAFRAME AND ADD PREFIX TO COLUMNS BASED ON NAME\n",
    "    print(perf_dir + modelfiles[0])\n",
    "    perf = pd.read_csv(perf_dir + modelfiles[0]).add_prefix(modelfiles[0][:-4] + \"_\")\n",
    "    # GET COLUMN NAME CONTAINING \"BASIN\"\n",
    "    fcol = [col for col in perf.columns if 'basin' in col]\n",
    "\n",
    "    # LOOP FOR EACH CSV FILE\n",
    "    for ii in range(1, len(modelfiles)):\n",
    "        print(perf_dir + modelfiles[ii])\n",
    "        # GET NEXT CSV FILE TO DEFINE DATAFRAME AND ADD PREFIX TO COLUMNS BASED ON NAME\n",
    "        currdf = pd.read_csv(perf_dir + modelfiles[ii]).add_prefix(modelfiles[ii][:-4] + \"_\")\n",
    "\n",
    "        # GET COLUMN NAME CONTAINING \"BASIN\"\n",
    "        basin_col= [col for col in currdf.columns if 'basin' in col]\n",
    "        # JOIN ON MATCHING BASINS\n",
    "        perf = perf.merge(currdf, how=\"inner\", left_on=fcol, right_on=basin_col)\n",
    "    \n",
    "    # GET COLUMN NAME CONTAINING \"FITMET\"\n",
    "    perf_met = [col for col in perf.columns if fitmet in col]\n",
    "\n",
    "    # CLEAN UP NONSENSICAL DATA (EG, BASIN LABELS)\n",
    "    # SO LETS GET A LIST OF VARIABLE NAMES WE WANT TO KEEP.\n",
    "\n",
    "    # TO START WE WILL KEEP THE SAME VARIABLES AS Kratzert et al. 2019, AS SHOWN BY OUR\n",
    "    # INTERNAL SPREADSHEET Attributes_CAMELS_vs_NHDPlus\n",
    "    varstokeep = ['organic_frac',\n",
    "    'elev_mean_x',\n",
    "    'slope_mean',\n",
    "    'area_gages2',\n",
    "    'soil_depth_pelletier',\n",
    "    'sand_frac',\n",
    "    'silt_frac',\n",
    "    'clay_frac',\n",
    "    'geol_permeability',\n",
    "    'p_mean',\n",
    "    'pet_mean',\n",
    "    'aridity',\n",
    "    'frac_snow',\n",
    "    'high_prec_freq',\n",
    "    'high_prec_dur',\n",
    "    'low_prec_freq',\n",
    "    'low_prec_dur']\n",
    "\n",
    "    camels_df = camels_df.merge(perf, how=\"inner\", left_on=\"hru_id\", right_on=fcol)\n",
    "    \n",
    "    inputdataset = camels_df[varstokeep]\n",
    "    outputdataset = camels_df[perf_met]\n",
    "    \n",
    "    \n",
    "    nsplits = 5\n",
    "    kf = KFold(n_splits=nsplits, shuffle=True)\n",
    "\n",
    "\n",
    "    testvalues = np.zeros((inputdataset.shape[0], outputdataset.shape[1]))                  # CONTAINER FOR PERFORMANCE VALUES WHEN BASIN IN TEST SET\n",
    "    modelno = 0                                                              # COUNTER FOR MODEL CONTAINER\n",
    "    test_modelno = np.zeros((inputdataset.shape[0], outputdataset.shape[1]))                # MODEL IN WHICH BASIN WAS IN TEST SET\n",
    "    test_modellist = list()                                                  # MODEL CONTAINER\n",
    "    featureimportance = list()\n",
    "    r2train = list()\n",
    "    r2test = list()\n",
    "    msetest = list()\n",
    "    msetrain = list()\n",
    "    msemeta = list()\n",
    "    trainlist_x = list()\n",
    "    trainlist_y = list()\n",
    "\n",
    "    currout=outputdataset\n",
    "\n",
    "    # KFOLD SPLIT OF DATASETS\n",
    "    for train, test in kf.split(inputdataset):            \n",
    "        # CODE FOR INDIVIDUAL MODEL TRAINING\n",
    "        for ii in range(currout.shape[1]):\n",
    "            # TRAIN MODEL ON TRAINING SET\n",
    "            model = None\n",
    "            model = ModelSelector(rf_kwargs=rf_kwargs)\n",
    "            model.fit(inputdataset.iloc[train, :], currout.iloc[train, ii])\n",
    "\n",
    "            # PERFORM PREDICTION ON TRAIN SET AND GET FIT METRICS\n",
    "            train_pred = model.predict(inputdataset.iloc[train, :])\n",
    "            trainrms = mse(train_pred, currout.iloc[train, ii].to_numpy())\n",
    "            trainr2 = r2_score(train_pred, currout.iloc[train, ii].to_numpy())\n",
    "\n",
    "            # PERFORM PREDICTION ON TEST SET AND GET FIT METRICS\n",
    "            model_pred = model.predict(inputdataset.iloc[test, :])\n",
    "            testrms = mse(model_pred, currout.iloc[test, ii].to_numpy())\n",
    "            testr2 = r2_score(model_pred, currout.iloc[test, ii].to_numpy())\n",
    "            \n",
    "            # GET FEATURE \n",
    "            fi = getFeatureImportance(model, test, inputdataset, currout.iloc[:,ii], modelno)\n",
    "            featureimportance.append(fi)\n",
    "\n",
    "            # SAVE VALUES IN CONTAINERS ABOVE\n",
    "            testvalues[test, ii] = model_pred\n",
    "            test_modelno[test, ii] = modelno\n",
    "            modelno = modelno + 1\n",
    "            test_modellist.append(model)\n",
    "            msetest.append(testrms)\n",
    "            msetrain.append(trainrms)\n",
    "            r2train.append(trainr2)\n",
    "            r2test.append(testr2)\n",
    "            trainlist_x.append(train_pred)\n",
    "            trainlist_y.append(outputdataset.iloc[train, :])\n",
    "\n",
    "            # print(f\"Test R2: {testr2:.3f} | Meta R2: {metar2:.3f} | Test MSE: {testrms:.3f} | Meta MSE: {metarms:.3f} | Model ID: {ii}\")\n",
    "\n",
    "    metrics = (r2train, r2test, msetrain, msetest)\n",
    "    trainlists = (trainlist_x, trainlist_y)\n",
    "    \n",
    "    return camels_df, inputdataset, outputdataset, testvalues, test_modelno, test_modellist, featureimportance, metrics, trainlists\n",
    "\n",
    "def softmax(x):\n",
    "    # ADAPTED FROM https://www.delftstack.com/howto/numpy/numpy-softmax/\n",
    "    maxx = np.max(x,axis=1,keepdims=True) #returns max of each row and keeps same dims\n",
    "    e_x = np.exp(x - maxx) #subtracts each row with its max value\n",
    "    sumx = np.sum(e_x,axis=1,keepdims=True) #returns sum of each row and keeps same dims\n",
    "    f_x = e_x / sumx \n",
    "    return f_x\n",
    "\n",
    "def ensemblePerf(perf, target, threshold=0.2, n=1, softmaxflag = False):\n",
    "    # TODO: CONSIDER DOING A MORE ROBUST THING WERE YOU CALCULATE THE DIFFERENCE\n",
    "    # BETWEEN EACH PERFORMANCE AND THE MINIMUM OF THE ROW WITHOUT ITSELF\n",
    "    # THEN, RUN A NP.WHERE TO ESTABLISH THAT IF BIGGER THAN THRESHOLD, SET TO NAN\n",
    "    # AND CALCULATE THE NORMALIZED PERFORMANCE FROM THERE, ONLY NECESSARY WITH MORE THAN\n",
    "    # 2 MODELS\n",
    "\n",
    "    # RANGE OF EACH ROW\n",
    "    rn = np.max(perf, axis=1) - np.min(perf, axis=1)\n",
    "    \n",
    "    # THIS IS THE BEST PREDICTED MODEL\n",
    "    # CALCULATED AS THE MAXIMUM OF THE PREDICTED PERFORMANCES\n",
    "    # BEST PRED MODEL = BPM\n",
    "    bpm_idx = np.argmax(perf, axis=1)\n",
    "    bpm_perf = perf[np.arange(perf.shape[0]), bpm_idx]\n",
    "\n",
    "    # CALCULATE THE VALUES FOR ENSEMBLING MODELS\n",
    "    \n",
    "    # EXPONENTIAL SETUP\n",
    "    if not softmaxflag:\n",
    "        perf_w = perf ** n\n",
    "        row_sums = perf_w.sum(axis=1)\n",
    "        perf_norm = perf_w / row_sums[:, np.newaxis]\n",
    "    else: \n",
    "        perf_norm = softmax(perf)\n",
    "    \n",
    "    ens_perf = np.sum(target * perf_norm, axis=1)\n",
    "\n",
    "    # THIS IS THE BEST ACTUAL MODEL\n",
    "    # CALCULATED AS THE MAXIMUM OF THE ACTUAL PERFORMANCES\n",
    "    # BEST ACTUAL MODEL = BAM\n",
    "    target_perf = target.to_numpy()\n",
    "    bam_idx = np.argmax(target_perf, axis=1)\n",
    "    bam_perf = target_perf[np.arange(perf.shape[0]), bam_idx]\n",
    "    \n",
    "    # CREATE OUTPUT STRUCTURE\n",
    "    output = np.where(rn < threshold, ens_perf, bam_perf)\n",
    "    return output, perf_norm\n",
    "\n",
    "def ensemblePerf(perf, target, threshold=0.2, n=1, softmaxflag = False):\n",
    "    # TODO: CONSIDER DOING A MORE ROBUST THING WERE YOU CALCULATE THE DIFFERENCE\n",
    "    # BETWEEN EACH PERFORMANCE AND THE MINIMUM OF THE ROW WITHOUT ITSELF\n",
    "    # THEN, RUN A NP.WHERE TO ESTABLISH THAT IF BIGGER THAN THRESHOLD, SET TO NAN\n",
    "    # AND CALCULATE THE NORMALIZED PERFORMANCE FROM THERE, ONLY NECESSARY WITH MORE THAN\n",
    "    # 2 MODELS\n",
    "\n",
    "    # RANGE OF EACH ROW\n",
    "    rn = np.max(perf, axis=1, keepdims=True) - perf\n",
    "\n",
    "    # CALCULATE THE VALUES FOR ENSEMBLING MODELS\n",
    "    \n",
    "    # EXPONENTIAL SETUP\n",
    "    if not softmaxflag:\n",
    "        perf_w = perf ** n\n",
    "        perf_w = np.where(rn > threshold, 0, perf_w)\n",
    "        row_sums = perf_w.sum(axis=1)\n",
    "        perf_norm = perf_w / row_sums[:, np.newaxis]\n",
    "    else: \n",
    "        perf = np.where(rn > threshold, 0, perf)\n",
    "        perf_norm = softmax(perf)\n",
    "    \n",
    "    n_weighting = np.count_nonzero(np.any(perf_norm == 1, axis=1))\n",
    "    n_partialweighting = np.count_nonzero(np.any(perf_norm == 0, axis=1))\n",
    "    n_total = perf_norm.shape[0]\n",
    "    print(f\"Not weighting in {n_weighting} of {n_total} basins, partial weighting in {n_partialweighting - n_weighting}\")\n",
    "    \n",
    "    return perf_norm\n",
    "\n",
    "\n",
    "\n",
    "def getLSTMTS(lstm_dir = '/home/ottersloth/neuralhydrology/nextgen-form-eval/run_hourly/runs/full_runs/runs/'):\n",
    "    file_list = os.listdir(lstm_dir)\n",
    "    file_list.remove('failed_runs')\n",
    "    file_list.remove('hydro_signatures')\n",
    "\n",
    "    num_files = len(file_list)\n",
    "\n",
    "    lstm_results = {}\n",
    "    lstm_results['basin_id'] = []\n",
    "    lstm_results['sim'] = []\n",
    "\n",
    "    for i in range(num_files): \n",
    "        lstm_test_dir = os.path.join(lstm_dir,file_list[i])\n",
    "        lstm_test_file = os.path.join(lstm_test_dir,\"test\", \"model_epoch003\",\"test_results.p\")\n",
    "        with open(lstm_test_file, \"rb\") as fp:\n",
    "            results = pickle.load(fp)\n",
    "\n",
    "        basin_id = [i for i in results.keys()]\n",
    "\n",
    "        for j in range(len(basin_id)):\n",
    "            sim = results[basin_id[j]]['1H']['xr']['QObs_CAMELS(mm/h)_sim'].values\n",
    "            sim = [float(sim[i]) for i in range(len(sim))]\n",
    "\n",
    "            lstm_results['sim'].append(sim)\n",
    "            lstm_results['basin_id'].append(basin_id[j])\n",
    "\n",
    "    df_lstm_results = pd.DataFrame(lstm_results)\n",
    "    return df_lstm_results\n",
    "\n",
    "def getWeightedFit(rfargs = {\"n_estimators\": 15}, weightparams={\"threshold\":0.2, \"n\":1, \"softmax\": False}):\n",
    "    \n",
    "    # RUN RF MODELS AND GET OUT OF BAG (OOB) \"TESTVALUES\"\n",
    "    camels_df, inputs, target, testvalues, test_modelno, test_modellist, featureimportance, metrics, trainlists = runFitMetric_getMSE(\"nnse\",rf_kwargs=rfargs)\n",
    "    outstruct_rf = (camels_df, inputs, target, testvalues, test_modelno, test_modellist, featureimportance, metrics, trainlists)\n",
    "    \n",
    "    # GET WEIGHTS FROM OOB TESTVALUES\n",
    "    weights = ensemblePerf(testvalues, target, **weightparams)\n",
    "    \n",
    "    # ADD WEIGHTS TO CAMELS AND DEFINE RUNNING DF \n",
    "    camels_df['weight_CFE'] = weights[:, 1]\n",
    "    camels_df['weight_LSTM'] = weights[:, 0]\n",
    "    camels_df['weight_NWM'] = weights[:, 2]\n",
    "    weights_df = camels_df[['hru_id', 'weight_CFE', 'weight_LSTM', 'weight_NWM']]\n",
    "    \n",
    "    # GET TIMSERIES OF LSTM RESULTS\n",
    "    df_lstm_results = getLSTMTS()\n",
    "    \n",
    "    # DEFINE DIR WITH ALL STREAMFLOW DATA \n",
    "    q_dir = \"/home/ottersloth/data/camels_hourly/usgs_streamflow\"\n",
    "\n",
    "    # TEST PERIOD\n",
    "    test_start=datetime.strptime(\"2002-09-30 23:00:00\", '%Y-%m-%d %H:%M:%S')\n",
    "    # ORIGINALLY UNTIL 11 PM, CHANGED BECAUSE COMPARISON FUNCTION IS INCLUSIVE LATER ON\n",
    "    test_end=datetime.strptime(\"2007-09-30 22:00:00\", '%Y-%m-%d %H:%M:%S') \n",
    "\n",
    "    # GET ALL CFE VALIDATION FILES \n",
    "    cfe_dir = \"/home/ottersloth/cfe_calibration/results/val_runs/\"\n",
    "    os.chdir(cfe_dir)\n",
    "    filelist = glob.glob(\"*.json\")\n",
    "    \n",
    "    # GET NWM FILES\n",
    "    nwmdir = \"/home/ottersloth/cfe_calibration/NWM_streamflow_results.csv\"\n",
    "    nwm = pd.read_csv(nwmdir)\n",
    "\n",
    "    basinlist = list()\n",
    "    outnnselist = list()\n",
    "\n",
    "\n",
    "    for i in range(len(filelist)): \n",
    "\n",
    "        # GET FILE NAME FROM CONTAINER\n",
    "        file = filelist[i]\n",
    "\n",
    "        # GET BASIN ID BY SPLITTING FILENAME\n",
    "        basinid_raw = file.split(\"_\")[0]\n",
    "        basinid = '%08d' % int(basinid_raw) # ZERO PADDING\n",
    "\n",
    "        # GET TIMESERIES FROM LSTM DATAFRAME\n",
    "        match = df_lstm_results[df_lstm_results[\"basin_id\"] == basinid]\n",
    "\n",
    "        # SINCE NOT THE SAME BASINS WERE RUN, CHECK WE ACTUALLY GOT A MATCH\n",
    "        if match.shape[0] != 1:\n",
    "            print(f\"Skipping {basinid}\")\n",
    "            continue\n",
    "\n",
    "        # CONVERT TIMESERIES TO NP ARRAY\n",
    "        ts_lstm = np.array(match.iloc[0, 1])\n",
    "\n",
    "        # NOW READ CFE FILE\n",
    "        with open(file) as json_file:\n",
    "            data = json.load(json_file)\n",
    "\n",
    "        # CONVERT TIMESERIES TO NUMPY ARRAY\n",
    "        ts_cfe = np.array(data['validation sims'])\n",
    "        \n",
    "        # GET NWM TIMESERIES\n",
    "        ts_nwm = nwm[basinid].to_numpy()\n",
    "\n",
    "        # GET MODEL WEIGHTS \n",
    "        weights = weights_df[weights_df[\"hru_id\"] == int(basinid)].to_numpy()\n",
    "        weights = weights[0][1:]\n",
    "\n",
    "        # weights = weights[1:]\n",
    "\n",
    "        # APPLY WEIGHTS\n",
    "        ts_final = (weights[0] * ts_cfe + weights[1] * ts_lstm  + weights[2] * ts_nwm)\n",
    "\n",
    "        # GET Q FROM USGS FOR TEST PERIOD\n",
    "        q_read = pd.read_csv(f\"{q_dir}/{basinid}-usgs-hourly.csv\")\n",
    "        q_read[\"datetime\"] = pd.to_datetime(q_read['date'], format='%Y-%m-%d %H:%M:%S') # CONVERT TO DATETIME\n",
    "        q_match = q_read[q_read.datetime.between(test_start, test_end)]\n",
    "        ts_q = q_match['QObs_CAMELS(mm/h)'].to_numpy()\n",
    "\n",
    "        # CALCULATE NSE \n",
    "        nse_calc = nse(ts_q, ts_final)\n",
    "\n",
    "        # CALCULATE NNSE\n",
    "        nnse = 1 / (2 - nse_calc)\n",
    "\n",
    "        # print(f\"{basinid} {nnse}\")\n",
    "        if i % 100 == 0:\n",
    "            print(f\"{i} / {len(filelist)}\")\n",
    "\n",
    "        # SAVE IN PROPER FORMAT\n",
    "        basinlist.append(basinid)\n",
    "        outnnselist.append(nnse)\n",
    "        \n",
    "    return basinlist, outnnselist, outstruct_rf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e7ec4642",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/ottersloth/ensemblennse/data/data/hourly_performances/LSTM.csv\n",
      "/home/ottersloth/ensemblennse/data/data/hourly_performances/CFE.csv\n",
      "/home/ottersloth/ensemblennse/data/data/hourly_performances/nwm.csv\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ottersloth/anaconda3/envs/ensemble/lib/python3.8/site-packages/geopandas/geodataframe.py:1483: FutureWarning: Passing 'suffixes' which cause duplicate columns {'gauge_id_x'} in the result is deprecated and will raise a MergeError in a future version.\n",
      "  result = DataFrame.merge(self, *args, **kwargs)\n",
      "/home/ottersloth/anaconda3/envs/ensemble/lib/python3.8/site-packages/geopandas/geodataframe.py:1483: FutureWarning: Passing 'suffixes' which cause duplicate columns {'gauge_id_x'} in the result is deprecated and will raise a MergeError in a future version.\n",
      "  result = DataFrame.merge(self, *args, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Not weighting in 42 of 495 basins, partial weighting in 200\n",
      "0 / 495\n",
      "100 / 495\n",
      "200 / 495\n",
      "300 / 495\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_171894/1159985010.py:339: DtypeWarning: Columns (3) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  q_read = pd.read_csv(f\"{q_dir}/{basinid}-usgs-hourly.csv\")\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "400 / 495\n"
     ]
    }
   ],
   "source": [
    "basins, nnse, outstruct = getWeightedFit(rfargs = {\"n_estimators\": 15}, weightparams={\"threshold\":0.2, \"n\":1, \"softmaxflag\": False})\n",
    "\n",
    "camels_df, inputdataset, outputdataset, testvalues, test_modelno, test_modellist, featureimportance, metrics, trainlists = outstruct"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "485d3c8b",
   "metadata": {},
   "outputs": [],
   "source": [
    "n_models = 3\n",
    "n_folds = 5\n",
    "topn = 4\n",
    "width = .4\n",
    "\n",
    "# -------------EXTREMELY IMPORTANT------------------------------------------- \n",
    "# DEFINE THE STARTING INDICES OF THESE LISTS AS THE SAME AS SHOWN IN THE CODE\n",
    "# BLOCK ABOVE (ie, FIRST COLUMN = 0)\n",
    "cfe_i = 1\n",
    "lstm_i = 0\n",
    "nwm_i = 2\n",
    "\n",
    "cfe_idx = np.arange(cfe_i, n_models*n_folds, n_models)\n",
    "lstm_idx = np.arange(lstm_i, n_models*n_folds, n_models)\n",
    "nwm_idx = np.arange(nwm_i, n_models*n_folds, n_models)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "934ff6bc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'NWM')"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams['figure.figsize'] = (15, 5)\n",
    "\n",
    "def getList(r, ax, c, topn=5):\n",
    "\n",
    "    features = featureimportance[0].shape[0] # NUMBER OF FEATURES\n",
    "    sum = np.zeros((features, ))\n",
    "    ss = np.zeros((features, ))\n",
    "    for i in r:\n",
    "        # POOL MEANS AND STANDARD DEVIATIONS:\n",
    "        # - GROUP MEAN IS MEAN OF ALL MEANS \n",
    "        # - GROUP STD IS DEFINED BY SQRT(SUM(STDs^2)), SINCE ALL SAMPLE SIZES ARE EQUAL\n",
    "        currdf = featureimportance[i]\n",
    "        sum = sum + np.abs(currdf[str(i) + \"_mean\"].to_numpy()).T\n",
    "        ss =  ss + currdf[str(i) + \"_std\"].to_numpy().T\n",
    "\n",
    "    means = sum / len(featureimportance)\n",
    "    means = means/means.sum() * 100 # NORMALIZE TO PERCENT\n",
    "    \n",
    "    std = np.sqrt(ss)  \n",
    "    \n",
    "    ind = np.argsort(means)[-topn:]\n",
    "    \n",
    "    labels = featureimportance[0].index.values[ind]\n",
    "    selmeans = means[ind]\n",
    "    \n",
    "    # REVERSE TO SHOW IN DESCENDING ORDER\n",
    "    #selmeans = selmeans[::-1]\n",
    "    #labels = labels[::-1]\n",
    "    \n",
    "    ax.barh(np.arange(selmeans.shape[0]), selmeans, color=c)\n",
    "    ax.set_yticks(np.arange(selmeans.shape[0]))\n",
    "    ax.set_yticklabels(labels, fontsize=12)\n",
    "    ax.set_xlim((0, 90))\n",
    "    ax.set_xlabel(\"Perc. Variable Importance\")\n",
    "    \n",
    "    return ax\n",
    "\n",
    "plt.rcParams['figure.figsize'] = (15, 5)\n",
    "figs, axs = plt.subplots(1,3, constrained_layout=True, facecolor=\"white\")\n",
    "\n",
    "\n",
    "figs.suptitle('Variable Importance via RFR Permutation (Top 5)', fontsize=18, x=0.55)\n",
    "axs[0] = getList(lstm_idx, axs[0], 'blue')\n",
    "axs[0].set_title(\"LSTM\")\n",
    "axs[1] = getList(cfe_idx, axs[1], 'red')\n",
    "axs[1].set_title(\"CFE\")\n",
    "axs[2] = getList(nwm_idx, axs[2], 'green')\n",
    "axs[2].set_title(\"NWM\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "b40ea52e",
   "metadata": {},
   "outputs": [],
   "source": [
    "lstm_trainr2  = np.array(metrics[0])[lstm_idx]\n",
    "lstm_testr2   = np.array(metrics[1])[lstm_idx]\n",
    "lstm_trainmse = np.array(metrics[2])[lstm_idx]\n",
    "lstm_testmse  = np.array(metrics[3])[lstm_idx]\n",
    "\n",
    "\n",
    "cfe_trainr2  = np.array(metrics[0])[cfe_idx]\n",
    "cfe_testr2   = np.array(metrics[1])[cfe_idx]\n",
    "cfe_trainmse = np.array(metrics[2])[cfe_idx]\n",
    "cfe_testmse  = np.array(metrics[3])[cfe_idx]\n",
    "\n",
    "nwm_trainr2  = np.array(metrics[0])[nwm_idx]\n",
    "nwm_testr2   = np.array(metrics[1])[nwm_idx]\n",
    "nwm_trainmse = np.array(metrics[2])[nwm_idx]\n",
    "nwm_testmse  = np.array(metrics[3])[nwm_idx]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "51ff793f",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(15, 396)\n",
      "(15, 396, 3)\n",
      "(1980,)\n",
      "(1980,)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABEAAAAFwCAYAAAC4vQ5FAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOydeXxM1/vH35N9mewrQhIUFWssVUullha1FEWVWCpp0UVUv5ZvK7H0q6LUUt9qJX5IrLVXFV9VQbWW0iCxL0EsIZFF9mTm/v4Yc80kk42EiPP2mpfMPeee+5zJ5DNnnvuc51FIkiQhEAgEAoFAIBAIBAKBQFCFMXrWBggEAoFAIBAIBAKBQCAQVDTCASIQCAQCgUAgEAgEAoGgyiMcIAKBQCAQCAQCgUAgEAiqPMIBIhAIBAKBQCAQCAQCgaDKIxwgAoFAIBAIBAKBQCAQCKo8wgEiEAgEAoFAIBAIBAKBoMojHCCCZ86KFStQKBRcunSpyD5qtZrly5fTunVrHB0dsba2pk6dOrz77rscPXoUAIVCUeLDy8sLgGnTpqFQKLC0tCQ1NbVIm0qyq6yMGDFCzx4rKyt8fX35v//7v0J9tTYaemhtioqK0jtuYmJCrVq1GDt2LMnJyeVmt0AgEBjir7/+YuDAgVSvXh0zMzOcnJzo2rUrK1euRKVSAYV1SvcRHh4uj1Wcdm/durXcbPby8tIb297enq5du/LHH38U6uvn52fQHg8PD7lPQa02NzenYcOGfPPNN6jV6nKzWyAQCAqiXa/a29sXWvfl5+ejUCiYNm0aoNFrhUJBRESEXj+VSoWNjQ0mJiY8ePBAr23Hjh0oFAp++eUXvespFAouXLhQyB5dvf/tt9/KbZ5l0VldGws+tDbFxcXpHTc2Nsbd3Z0hQ4Zw48aNcrNbUDkxedYGCASl4fPPP2fRokUEBQUxbdo0TE1NuXDhAlu2bOHIkSO0bt2av/76S++cvn370rRpU1n4AczNzfX6mJqasnHjRkaNGqV3PCIiAhsbm0IfBOWBi4sLP//8MwAJCQksXLiQUaNGYWtryzvvvFOo/x9//IGxsbHesZo1a+o9X7RoEa1atSIzM5O9e/cSGhrKjRs32L59e7nbLxAIBAALFizgs88+o1OnToSGhuLp6UlycjL/+9//GDNmDPb29vTp00fur9UpXerUqaP3fMSIEXz44YeFrlW/fv1ytf3NN99k2rRpqNVqLl68yPTp0+nRowenTp2SHeVamjRpwo8//qh3rOBnCTzS6vv377NixQomTpyIkZEREyZMKFfbBQKBoCCpqamEhoYye/bsIvu0bNkSKysrDhw4wLBhw+TjJ06cICMjAwsLCw4dOkS3bt3ktgMHDmBkZET79u31xrKxsSEyMpKZM2fqHa/I9TOUTWc3bNig56wGaNiwod7zKVOm0Lt3b3Jzczl8+DDTp0/n7NmzHDlyBFNT0wqZg+DZIxwggkpPVlYW//3vf/nkk0+YO3eufLxr16589NFHsue3TZs2eueZm5vj7Oxc6Lgu/fr1IzIyUs8BcuPGDfbv38+wYcNYsWJF+U4GMDMz07Opc+fO1KxZk/DwcIMOkFdeeQUTk+L/VF9++WV5zE6dOnH37l3Cw8O5c+cO7u7u5TsBgUDwwnPgwAE+++wzPv74YxYtWqTX1qdPHz777DMyMjL0juvqVFHUqFGjxD7lge5nQ9u2balbty7t27dn3bp1TJ48Wa+vjY1NqWzS1epu3bpx6tQpwsLChANEIBBUOG+88QbfffcdQUFBRa77TE1NadOmDQcOHNA7fuDAAXx8fHBzc+PAgQOFHCBNmjTB3t5e75x+/fqxatUqZsyYgUKhADTr9U2bNtG/f/8KWT9D2XS2WbNm1K1bt9jxateuLev7a6+9Rl5eHl9++SXHjx9/Kp9FgmeD2AIjqPRkZGSQm5tbpKAbGT3+23jYsGEcOHCAa9euycciIyOpVasWr7322mOPWxaUSiX16tXj+vXr5Tamr68vQLmOKRAIBFpmz56No6Mjc+bMMdhep04dmjRp8pStenzKWzONjIxo2rSp0GCBQPBU+PLLLwH4z3/+U2y/jh07cvHiRe7cuSMfO3DgAB06dKB9+/Z6zpHMzEyOHz9ucD3s7+/PtWvX9LYObtmyBZVKRf/+/Z90OqWiInRWrJ9fDIQDRFDpcXZ2xtvbm7lz5/LDDz+Uqyh16NABLy8vVq9eLR+LjIxk6NChske7olGpVNy4caNQKLhue35+vvwozZ7yuLg4jI2NC4VyCwQCwZOiUqmIiorijTfewMLCotTnqdVqPS3T5gjRRZIkvT7aR0UTFxcHFN6So6WgPZIklWrMosYTCASC8qRatWp8/PHHLF26VO+mXkG0zgyto0OSJP744w86dOhAhw4dOHbsGNnZ2YAmZ0heXp5BB4inpyevvfYakZGR8rGIiAj69u2LUqksz6kVS3E6W3D9bOgzx9B4UPRngaBqIBwggueCNWvWYG1tzZgxY/D09KRGjRqMGjVKToD6uCgUCoYOHSoL+NGjRzl37pze3siKQCvGt27dYty4caSmpsre+4JYWFhgamoqPwzZpv1i8eDBA7Zu3cqSJUsICgrC1dW1QuchEAhePBITE8nKysLT07NM57355pt6Wmbo/FmzZun10T4SExPLy3zgkaMlNzeXM2fOMHr0aF566SXef//9Qn0PHTpUyJ5ly5YV6qddbN+7d4+vv/6a48ePF9ofLxAIBBXFpEmTsLS0ZPr06UX2adOmDWZmZrIDJCYmhvv379OhQwfatGmDWq3m8OHDwCMnSVER0cOGDWPDhg1kZ2dz+/ZtfvvttwpfP5dFZxs0aKCn2x07dizUR7t+zszM5Pfff+c///kP/fv3p0WLFhU6D8GzReQAETwXtGnThvPnz7Nv3z5+++03/vrrL1auXMmKFStYvnz5EwnusGHDmDlzJseOHSMiIoI2bdrw0ksvcejQoRLPlSRJz6OszSRdHDdv3tRLrKRQKFi7di2vvPKKwf6HDx/WG9PJyalQnzfffFPv+VtvvcU333xTov0CgUDwtPjvf/9L69at5edmZmaF+rz//vuMGTOm0PGC+891UavVepFxRkZGJW6NXLNmDWvWrJGfOzo68ueff+Lg4FCob9OmTfWq1QAGo+sKRsPMmTOHt99+u1g7BAKBoLxwdHRkwoQJTJ8+nUmTJhmMYrCwsKB169ayc+PAgQN4eXnJyUKbN2/OgQMH8PPz48CBA7z88su4uLgYvN6AAQP4+OOP2b59O9euXcPd3Z3OnTsXyjFiiILrZ6DEfHda+3UpTme3bNmilwTVxsamUJ8PP/xQL/G2r68vq1atKtEOwfONiAARPDeYm5vTrVs35s6dy6FDhzhz5gzu7u589tlnTzRu3bp1efXVV1m2bBnr1q0rkzNl5cqVet7l0oTMubq6cuzYMY4cOcKaNWvw9vbm/fff59y5cwb7t2jRgpYtW8oPb2/vQn3++9//cuzYMX777TcGDRrEjh07xJ1HgUBQITg5OWFpaVlsmLUh6tWrp6dlhnKEVKtWTa+P9lHcwvj999/X02FDURwF6d69O8eOHePPP/9kwYIFZGVl0a9fPzn0WxelUlnIHmdn50L9Dh8+zNGjR9myZQu+vr5MnjyZqKioEm0RCASC8mL8+PE4OjoSHBxcZJ/XXnuNmJgYkpOT5fwfWjp06MCBAwfIzc3lyJEjxebDs7Gx4e233yYyMpKIiAiGDBlS6rx8+/fvLxRZVxrKorONGjXS021D1cS+/PJLjh07xv79+/n44485ceIEY8eOLZUtgucXEQEieG6pV68egwYNYv78+dy9e/eJtnsMGzaMjz76CBMTEwYNGlTq83r16sWxY8fk54ZKIxbE1NSUli1bAtC6dWt8fX1p0qQJEyZMYMeOHWU3nkdfLEBTBSYhIYFZs2YxcuTIQiVzBQKB4EkwMTHBz8+PPXv2kJOTUyrdq0imTZvGxx9/LD835JwoiKOjo6yZr776KnZ2dowcOZLvvvuOf/3rX49lR4sWLTAxMaFVq1Z06NCB+vXr88knn3Dy5MknStYtEAgEpUWpVDJlyhQmTJhQpJa99tprzJo1i4MHD3Lw4EG9LTMdOnTghx9+4M8//yQrK8vgthFdhg0bxltvvYVarWbt2rWltrNFixZ66+eynFeeOuvp6Sl/Frz22ms8ePCA5cuXM3r0aL2IRUHVQnwiCyo9eXl5JCQkGGw7d+4clpaW2NnZPdE1Bg0aRO/evZk8eTKOjo6lPs/JyUnPu9y4ceMyX7t+/fp89NFH/Prrr4/1YVAQhULBggULyM3NLbYevEAgEDwukydPJikpqcgF9tWrVzl16tRTscXLy0tPhx8n+fPw4cPx9fXlm2++ITMz84ltcnJyIjg4mJiYGDZt2vTE4wkEAkFpGTt2LDVq1Cgyt1y7du0wMTFh2bJl3L59Wy8CpH379mRmZjJ//nyg6PwfWrp27crAgQMZPXo0Pj4+pbbRxsamUGRdWakInZ09e3aJeVQEzz/CASKoNOzatYuNGzfqPfbs2UNqaiq1atXC39+ftWvXcvDgQbZu3cq7777Lzp07+eSTT574DqSDgwNbtmx5ZoI3efJkrK2tmTFjRrmM17RpU/r378+yZcu4detWuYwpEAgEWl577TW+/fZbFi9eTNeuXVm9ejUHDx7k559/Zty4cTRq1IirV6+WedybN29y+PDhQo/bt29XwCweoVAomDFjBvfu3WPJkiXlMuaHH36Ih4cHX331VamqxggEAkF5YG5uTnBwMLt37zbYrlQqad68Odu3b8fFxYUGDRrIbc7OzjRo0IDt27dTu3ZtatSoUey1jI2NWbt2bbnpZlkpb511d3eXb0oeP368HCwUVEaEA0RQafjkk08YMGCA3mP8+PHY2try1VdfkZCQwL/+9S86d+7MsGHDuHbtGj/++GOViHJwdXXl008/5ZdffuGff/4plzFnzJhBXl4eoaGh5TKeQCAQ6BIUFMQff/yBvb09n3/+OZ06dWLEiBGcPXuWH3/8kV69epV5zBUrVvDqq68WeuiWKq8o3nrrLdq2bcvcuXPJysp64vHMzc2ZOnUqp06dYuvWrU9uoEAgEJSSkSNH8tJLLxXZ3rFjRyRJon379oXaOnTogCRJJUZ/VAYqQmcnT56MjY1Nud2UFFQ+FJK4LSEQCAQCgUAgEAgEAoGgiiMiQAQCgUAgEAgEAoFAIBBUeYQDRCAQCAQCgUAgEAgEAkGVRzhABAKBQCAQCAQCgUAgEFR5hANEIBAIBAKBQCAQCAQCQZVHOEAEAoFAIBAIBAKBQCAQVHlMnrUB5YGzszNeXl4VNn5GRgbW1tYVNn5lQsy1aiLm+nwSFxdHYmLiszbjiREaXX6IuVZNxFyfT6qCRlekPlel33VJiLlWTcRcn1+K0+cq4QDx8vLi77//rrDxo6Ki8PPzq7DxKxNirlUTMdfnk5YtWz5rE8oFodHlh5hr1UTM9fmkKmh0RepzVfpdl4SYa9VEzPX5pTh9FltgBAKBQCAQCAQCgUAgEFR5hANEIBAIBAKBQCAQCAQCQZVHOEAEAoFAIBAIBAKBQCAQVHmEA0QgEAgEAoFAIBAIBAJBlUc4QAQCgUAgEAgEAoFAIBBUeYQDRCAQCAQCgUAgEAgEAkGVRzhABAKBQCAQCAQCgUAgEFR5hANEIBAIBAKBQCAQCAQCQZVHOEAEAoFAIBAIBAKBQCAQVHlMnrUBz4IFCxaQkpJS6v7169dn2rRpFWZPZULMtXywt7cnKCioQsYWCKoyZdVnELpVVamouQp9Fggen7JotNCrqklFz1VotKCieSEdICkpKWX6w42KimLw4MEVZ1AlQsy1fHhRPgQFJZOVlcXVq1dp2LDhszbluaCs+gxCt6oqFTVXoc8CXU6dOsXLL7+MqanpszbluaAsGi30qmpS0XMVGi3Qcu/ePbKysqhVq1a5jiu2wAgEAkEFkZGRwYoVK9i8eTMPHjx41uYIBAKB4CGSJBEVFcXmzZs5evToszZHIBAIBDrcuXOHFStWsG3bNtRqdbmO/UJGgAgEAkFF8+DBAyIiIkhJSWHw4MHY2Ng8a5MEAoFAgMb58dtvv3Ho0CGaNWvGq6+++qxNEggEAsFDbt68yapVqzA1NeW9997DyKh8YzZEBIigXPjrr7949dVX6dixI4MHDyYvL+9ZmyQQPDNSU1NZvnw5qampDBkyhDp16jxrkwQvMEKfBYJHSJLErl27OHToEK1ataJPnz7lvrgWCMqC0GiB4BHXr18nIiICCwsLRo4ciZOTU7lfQyi+oFzw9PTk999/Z//+/dSuXZtt27Y9a5MEgmfC/fv3Wb58ORkZGfj7++Pl5fWsTRK84Ah9Fgg0qNVqtm/fzpEjR3j11Vfp0aMHCoXiWZsleMERGi0QaLh69SqRkZEolUpGjhyJg4NDhVxHOEBeIHx8fIiKiqqQsatXr46lpSUAJiYm4m6K4IUkMTGR5cuXk5OTw/Dhw6lZs+azNknwnCD0WSCoWNRqNVu3buXEiRN06NCBN954Qzg/BKVGaLRAULFcunSJ1atX4+DgwMiRI7G1ta2wa4m/MB28vLwwMzMjMTFR7/iZM2dQKBTExcU9NVuUSqX8MDIywtLSUn6+evXqxxozNjYWPz+/Mp+XnJyMQqFAqVRiZWWFp6cny5YtM9j36tWr7Ny5k549e5Zq7Pv379O3b1+sra3x9PRkzZo1RfaNi4ujR48eODg44O7uzscff0x+fr7c7ufnh4WFhfw61a9fX24bOnQo1apVw9bWlnr16hEeHl7K2QsEpSMhIYHly5ejVqsZMWIE1atXf9YmVSmK0udmzZpx/Phxoc/PWJ8Bzp49S6dOnbCzs6Nu3bps2bJFblu8eDEtW7bE3NycESNGFDpXaLSgIlGpVGzcuJFTp07RqVMnOnfuLJwf5UxxGi3W0JVbo3Nychg1ahSenp7Y2NjQvHlzdu7cWeS1Tp8+XeK1BIKycO7cOdauXYuLiwsjRoxAqVRW6PWEA6QA3t7erF27Vn5++vRpJEl66nakp6fLj1q1arF9+3b5+ZAhQ/T66joBKoLo6GicnZ1JT08nMzOTr7/+mg8//LDQh1xaWhrDhw8nMjISMzOzUo390UcfYWZmRkJCAqtXr2bMmDHExsYa7Dt27FhcXV25ffs20dHR7N+/n++//16vz+LFi+XX6fz58/LxKVOmEBcXR1paGj///DNffvklmZmZZXwlBALD3Lp1ixUrVmBkZMTIkSNxc3N71iZVSQzpc1ZW1lO3Q+hzYfLz8+nTpw89e/bk/v37LF26lKFDh3LhwgVAc4fzyy+/5P333zd4vtBoQUWRn5/PTz/9xJkzZ3jjjTd47bXXnrVJVZaCGp2VlSU0+jnQ6Pz8fGrWrMn+/ftJTU1l5syZDBw4UM9ppXstb2/vYq8lEJSFmJgYfvrpJ6pVq8awYcOwsrKq8GsKB0gB/P39iYiIkJ+vXLmyUPKVW7du0b9/f1xcXPD29mbRokVy2+zZs6lTpw42NjY0bNhQ7w4YaDzkc+fOpUmTJtjZ2TFo0CCys7PLbKeXlxehoaE0adIEa2tr8vPzS3Xt3377rcx2REdH4+vrKz/v2LEjKpWK5ORk+Vh+fj6DBw9m2rRpepEXxZGRkcGmTZuYOXMmSqWS9u3b07t3byIjIw32v3r1KgMHDsTCwgJ3d3e6detWavH18fHB3NwcAIVCgUKhICcnp1TnCgTFcePGDSIiIjA3N2fkyJE4Ozs/a5OqLIb0ediwYXp9hD4/G30+d+4ct27dYvz48RgbG9OpUyfatWsn9+/Xrx9vv/12kcnMhEYLKoK8vDzWrl3L+fPneeutt2jbtu2zNqlKU1Cjk5KSXhiNLupm6fOg0dbW1kybNg0vLy+MjIzo2bMn3t7eHD9+3OC1lEplsdcSCEpLdHQ0mzZtombNmvj7+8tbwSoa4QApQJs2bUhLS+Ps2bOoVCrWr1+Po6Oj3K5Wq+nVqxdNmzbl5s2b7N27lwULFrB7924A6tSpw8GDB0lNTSUkJIShQ4dy+/ZtvWv89NNP7Nq1i6tXr3Lq1ClWrFjxWLauXbuWHTt2kJKSgomJSamu/Th2/PPPP7Ro0QKAlJQUpkyZQosWLahbt66eLUeOHGHGjBn4+fmxfv16QBO1MXbsWIPjXrhwAWNjY+rVqycfa9q0aZFOjXHjxrFu3ToyMzO5efMmO3fupFu3bnp9pkyZgrOzM+3atSu0V3Ps2LFYWVnRoEEDqlWrhp2dXZGvjUBQGuLi4oiMjMTKyoqRI0fqaYWg/DGkz0OHDpXbhT4/O302tPiXJImYmJgi51gQodGC8iQnJ4fVq1dz5coV+vTpQ6tWrZ61SVWeghqdnJz8wmh0wYgOLc+jRickJHDhwgV8fHwe61oCQWn4+++/2bp1K97e3gwdOlS+CfI0EA4QA2g92Hv27KFBgwZ6oWjHjh3j3r17BAcHY2ZmRu3atQkMDGTdunUADBgwgOrVq2NkZMSgQYN46aWXOHr0qN74n376KdWrV8fR0ZFevXoRHR39WHZ++umn1KxZU/aWlebaj2NHdHQ0CxcuxNbWFgcHB+7evcuuXbv09s/6+/uTmJhIVFQUUVFRDBo0CIDvv/++0DYVLenp6YUWuHZ2djx48MBg/44dOxIbG4utrS0eHh60bNmSt99+W24PDQ3lypUr3Lx5kw8++IBevXpx+fJluf3777/nwYMHHDx4kH79+on9v4In4vLly6xevRo7OztGjhwpvqw9JQrqc40aNeQ2oc/PTp8bNGiAq6sr33zzDXl5efzvf/9j//79ZdrGIjRaUF5kZ2ezatUqrl+/Tr9+/WjevPmzNumFQVejLSwsXhiNLmqrz/Om0Xl5eQwZMoThw4fToEGDx7qWQFAShw8f5pdffqFevXq89957mJqaPtXrCweIAfz9/VmzZg0rVqwoFLp37do1bt26hb29vfyYNWsWCQkJAERERNCsWTO5LSYmppBX2N3dXf7ZysqK9PT0x7KzYIWJ0ly7rHbk5ORw9uxZTp06RVpaGhs3buTw4cPl8kZVKpWkpaXpHUtLS8PGxqZQX7VazZtvvkm/fv3IyMggMTGR5ORkJk2aJPd55ZVXsLGxwdzcnOHDh9OuXTt+/fVXvXGMjY1p37498fHx3Lt374nnIHgxOX/+PGvWrMHR0ZERI0YYfM8KKgahz4+oLPoMYGpqytatW9mxYwfu7u7MmzePgQMH4uHhUabrCo0WPCmZmZmsXLmSW7duMWDAABo3bvysTXqh0NXoglveqrJGq1SqQn2eN41Wq9X4+/tjZmbG4sWLH/taAkFxHDx4kF27dtGwYUMGDRqEiYnJU7dBOEAM4Onpibe3N7/++iv9+vXTa6tZsybe3t6kpKTIjwcPHvDrr79y7do1AgMDWbx4MUlJSaSkpNCoUaMKS6Kq6z2uqGvHxMRgbm5O7dq1Aejfvz+1atVi06ZNTzQuQL169cjPz+fixYvysZMnT8ohd7rcv3+fGzdu8PHHH2Nubo6TkxMjR44s5ODQRaFQFDn//Px8sb9c8FjExsayfv163NzcGDFiBNbW1s/apBcKoc+PqCz6rKVJkybs37+fpKQkdu/ezZUrV2jduvVjXV9otOBxSE9PZ8WKFdy7d49Bgwbx8ssvP2uTXjh0Ndre3l6vTWh05dVoSZIYNWoUCQkJbNq0Sc9J8zjXEggKIkkS+/btY+/evTRu3Jh33nkHY2PjZ2KLcIAUwbJly/j9998Lfblp3bo1tra2hIaGkpWVhUqlIiYmhmPHjpGRkYFCocDFxQWA5cuXl2n/85NQUdf+559/aNSokd4HRY8ePfj555+feGxra2v69etHcHAwGRkZHDp0iG3btuHv71+or7OzM97e3ixZsoT8/HxSUlJYuXIlTZs2BTT7Knfv3k12djb5+fmsXr2aAwcO8Oabb3L37l3WrVtHeno6KpWK3bt3s3bt2gqtLy0oG2FhYdSsWZOwsLBnbUqxnDp1io0bN1KjRg2GDRv21JI1CfQR+qyhsuizllOnTpGdnU1mZiZz587l9u3bcsnb/Px8srOzUalUqFQqWasBodFVlNi7sQzbPIxhW4YRezdWPhafFi8/1+07ftf4QsfLQlpaGitWrCA5OZn33ntPL1+B4Omi1WgjI/2vGUKjK69GjxkzhrNnz7J9+/ZCa5uC18rIyCjxWoKKobTr5fLQ1LKMY0jvdZEkiT179rB//358fX3p27dvIX14mggHSBHUqVOHli1bFjpubGzM9u3biY6OxtvbG2dnZwICAkhNTaVhw4ZMmDCBV199FTc3N06fPk27du2eir0Vde3o6GiaNGmid6xbt27s2bOnVJm3R48ezejRo4ts//7778nKysLV1ZXBgwezZMkSPY9y9+7dmTVrFgCbN29m165duLi4ULduXUxMTJg/fz6g2bP45Zdf4uLigrOzM9999x1bt26lfv36KBQKlixZgoeHBw4ODnz++ecsWLBA5GyoRMyYMYP4+HhmzJjxrE0pkuPHj7Nlyxa8vLzw9/fHwsLiWZv0wiL0WUNl0meAyMhIqlWrhqurK3v37mXPnj1yUrOvvvoKS0tLZs+ezapVq7C0tOSrr74CEBpdRQk/Ec6GMxvYcGYD4SfC5WNJWUnyc92+q06vKnS8tKSkpLB8+XLS0tIYOnSofMdd8GwQGq3hedHoa9eu8eOPPxIdHY27u7tc6WX16tUGr3XlypVC1xI8HUq7Xn5STS3rOIb0XoskSezcuZM///yT1q1b06tXr2fq/NAa9dzTokWLMvUPCQkpU/99+/aVqf/zjJhr+VDW91hFU9l/r0uXLpU8PDykpUuXPlEfSaqYuR4+fFgKCQmRIiMjpdzc3HIfvyjKqm2VlbLM43H+dir7+7s8EXN9ciqbPktS1fy9xiTESP6b/CX/zf5STEKMfCxyW6T8XLdv0M6gQse1bb3W9JIafNdAWnR4UaF+SUlJ0rfffit9/fXX0o0bNyp2UgWoChpdkWvoqvi+Lgox17JR3JquMml0Zfu96r5uJemmbltJzyXJ8Fy3nt0qNf+hubT17NZi7TKk95IkSSqVStq2bZsUEhIi7d69W1Kr1Y8z7ceiOG17+llHBAKBoACBgYEEBgYW20fX611S3/Lk0KFDcsWRd95555kkaxIIBILnDR9XHyL6RRQ6ds/2Hj6uPoWOz+823+A44SfC+fXir6gkFTMPzERCk5dhfrf53Lt3j4iICFQqFcOHD6datWoVMxmBQFCuPKs13fOO7np5/K7xrDq9CqCQfhbUVG0kh7ZvwedFERUXxY20G0TFRdGnQZ8i+xnSe7VazZYtWzh9+jQdO3bEz8+v0lR3E1tgBAJBhVDeeT2Cg4Px8PAgODj4scYu6zmSJBEVFcWePXto1KgRAwYMEM4PgUAgKCcK7i3XfR57N5bea3pTZ2EdfrvyG0YKzXLVztyO7nW6E+AbwJ07d1ixYgWSJDFixAjh/BAIniN013SCxyPAN4ChjYcS4BtQ5r6lPbcs19BFpVKxceNGTp8+TefOnXn99dcrjfMDhANEIBBUEOWd1yMwMJAbN24QGBhYprG1jo9JkyaV+hxJkti7dy9RUVE0a9aMfv36PbNM1QKBQFAVKbi3PPxEOCtOrmDAhgG889M77Li4gyspV4i5F0OeOg+Ay8mXQQELDi5gzvI5GBsbM3LkSFxdXZ/lVAQCQRnRXdMJHlGWm3XaKI+CEXWl6evj6kOAbwDhJ8LZdm4b43eNJzU7tVDC05KuYShJan5+PuvXr+fMmTN069aNDh06lGjf00Y4QAQCQYVQkd79soytdZZIklSqcyRJYteuXfzxxx+0bNmSPn36PPtkTQKBQFDFMHRH0tvemyvJV7iScgVPO09q29fGzcoNABNM6OjZkYz0DNbErOE4xxk5ciROTk7PchoCgUBQakpycDzpzcOiqm8N2zKMYZv1K7RondDT909n1elV3Eq/VebEqQUd2bm5uaxZs4YLFy7Qs2dP2rRp81jzqGhEPLdAIKgQSpPX42mMHRwczIwZMwgODi7xHEmS+OWXXzh+/Dht2rThzTffrFQhewKBQFBVKLhH3cfVh8i+kYQeCgUJJrWfhI+rD7F3Ywk/EU6AbwBWGVbMXz2f1uatmT1wNg4ODs9wBgKBQFA2Ssp9ortmfRy01VhQgJOlk5zvY8OZDSCBk5WTrLta57Oflx9RcVFUp3qZt7voOrBzcnJYs2YN169f5+2336ZZs2aPNYengXCACASCSkdYWJjeB8CkSZOQJIk5c+YQGBio116UUyP2biyhf4SCM+w6vgvQJIzSCn2Ab4BeSJ9arWbbtm2cPHmSDh060KlTJ+H8EAgEVYbS6GZ5o+u80Doz4tPi+e7Idyw5tgQ3pRs25jaazhLwUHLtze35/ervDNgwAKWJkvT8dOqb1yfxUCINHBswa9gslErlU5mDQCAQlBclOTie9OZhgG8ASZlJoNB3TiRlJYGEnnND1wndp0EfoqKimO9XdEJUQ2jHyM7OJjIyklu3btG/f38aNWr02HN4GggHiEAgqHQUDAFMTk6WjxfMAVLUB0VBLzhAWHgY/937X8w6m0HAo8zXKpWKzZs3ExsbS6dOnXjttdeewiwFAoHg6fEsqi4YqjzgleXF9KjpJGcnczbpLMaKh/mVdBwgAJGnI/XGmn5gOiFuIQwbNgwrK6unYr9AIBCUJ0/i4CiNE7uo6lsRffWP6TqnQaPVb5q8KbdvO7eN6funE9IxpNjqLwCZmZlERkZy9+5dBg4cSIMGDR5nek8VsbFdUGb++usvXn31VTp27MjgwYPJy8t71iYJKP+qK88S3Rwf2rJZlpaWsse8NDlAAnwDGNBwAAMaDiDAN4AA3wAU+xXkpeSh2K+QRT8/P5+ffvqJ2NhY3njjDeH8EDzXCH0WFEVRuqlNYqdNhGeoKktRlNQnwDeA7nW6k5SVROzdWAJ8A3CydKKpW1OMMMLS2BJ3pTsdanWgx0s95IfRw+WpAgWN7BrhjDP9HPoxfPhw4fwQPNcIjRY8LuVZXEA3d4f258TMRLl9+v7pRN+JZvr+6cWOk56ezooVK7h37x6DBw9+LpwfIBwggsfA09OT33//nf3791O7dm22bdv2rE0SUP5VVyqSkpw1utnBo6KikCQJJyenEre/hIWFcfr0acLCwmQveETfCHxcffBx9eHbWd/i4eHBt7O+5c9tf+Lh4cEHH3zA+fPn6dGjB23btn0a0xcIKgyhz4KiKKrqQsFEeOEnwom9G4v/Fn9WnFxRbEI8beUW/y3+Bp0gPq4+oIANZzYwZe8Uwk+E42zlzOIei6nvXJ9sVTY3H9zk7L2zxNyL4UHOAwBszW0BcDB1oFFqI5ormxPQLwBLS8tyfEUEgqeP0GjB41KexQV0k1Brf3a2cpbbQzqG0My9GSEdQ4ocIy0tjeXLl5OSksKQIUOoW7fuE9v1tBAOEEGZqV69urwIMTExERUyKgmVpaZ6aSJRyuKsKTiv4s6dNGkSubm5TJo0yeBYul8Apk+fzs2bN9m0aRN9+vShdevWpZyhQFB5EfosKCvaxW9IxxB5QRx+IpyrKVfxtveWo+UKRnvE3o0lKSsJB3MHYu/FMmXvFMPRIJLmcTHponyX0cfVh5bVWiIhAZCQmcDVlKtEXYvilwu/kJSdBEBqbip/mv7JP6p/iIjRD+EWCJ5HhEYLHpfSlg7WXYcXFaWnLYMbeiiU0D9CCfANwMLEQm7v06APJz48QV3HugbPT0lJYfny5aSnpzN06FC8vb3Lb6JPAfFXp4OXlxdmZmYkJibqHT9z5gwKhYK4uLinZotSqZQfRkZGWFpays9Xr1792ON6eXnx22+/FdmuUqlQKBQolUqsrKzw9PRk2bJlBvtevXqVnTt30rNnz8e25/79+/Tt2xdra2s8PT1Zs2ZNkX0XL15My5YtMTc3Z8SIEYXadV8zpVKJsbExn3zySanOrQpUlprqJTk3wsLCyMjIwN7eXnZqFOc0KTgvXYdIQWHPV+cDmpwh2rG0Y4+dMRbfH33Zdm4b2dnZdOrUCVtbW4KCgmjevHm5vw6C8qUofW7WrBnHjx9/IfQ5OTmZ48ePV4g+l0WLdbl48SIWFhYMHTq01GMNHTqUatWqYWtrS7169QgPL33JPUH5o01i16dBH+Z3my8vjEc0HUFk30g5WbRuyHRYWBhtfNqwefVmLEwtMFIYcSnpEqtOryL0j1BZl7Xa3LVOV15yeolXPV4lX51P7N1YWtVohamRaSF7tE4RY8mYlyxeorFnY7q/1L1MlQkET5/iNPpFWUMnJydX2Bq6vDQ6JyeHUaNG4enpiY2NDc2bN2fnzp1655w9e5ZOnTphZ2dH3bp12bJlS6muVRmpbNvDy8ueguvfbee2yWvcqdOmEh8fz9RpU4uN0gs/Ec5PsT+x5vQaTdEAA+NqdV9X15OSkvi///s/srOzGTZsGLVq1XqiuTwLhAOkAN7e3qxdu1Z+fvr0aSRJeup2pKeny49atWqxfft2+fmQIUMq7LqZmZk4OzuTnp5OZmYmX3/9NR9++GGhD7S0tDSGDx9OZGQkZmZmj329jz76CDMzMxISEli9ejVjxowhNtbwXuLq1avz5Zdf8v777xts133NEhISsLS0ZMCAAaU6V1A0ZRXrkiJRZsyYQXJyMkqlUnZqlCYiRGuH9hozZsxg3H/G6dUfbz38URSHdizt2OELwom+E83HMz+mevXqpKam8tdffzF9evH7GwWVB0P6nJWV9dTteFb6HB0djYmJSYXoc1m0uOB5rVq1KtNYU6ZMIS4ujrS0NH7++We+/PJLjh8/Xio7BU8HrVNEt1KWn5cfLpYuRN+JJmhKEOmJ6Sj2K/i689eMbjGar7t8zdDGQ0GBZsF8KBT/Lf5sv7id+LR4Dl4/yOH4w6RkpxD6Ryhf/v4leeoi8h9IoFaosbS35MitIzhZOunZIqicFNTorKysF06jK2oNXV4anZ+fT82aNdm/fz+pqanMnDmTgQMHyg6q/Px8+vTpQ8+ePbl//z5Lly5l6NChXLhwofQvRCWism0PLy97dB3SoJ+zo9E7jTCyM6LRO40I8A3A296bqylXC21nDPANoLZDbYyNjeUE1AXH1UYIanX9u0PfsXz5clQqFcOHD6dGjRpPNI9nhXCAFMDf35+IiEdhlitXrsTJyUmvj7bEj4uLC97e3ixatEhumz17NnXq1MHGxoaGDRsW8pp6eXkxd+5cmjRpgp2dHYMGDSI7O7vMdhZnA0BoaCg1atTAxsaG+vXrs3fvXvz9/bl+/Tq9evVCqVQyZ86cQuNmZWXh6+srP+/YsSMqlUquwgEacRw8eDDTpk2jfv36ZbZdS0ZGBps2bWLmzJkolUrat29P7969iYyMNNi/X79+vP3224V+H4bYuHEjrq6udOjQocznCvQpq1gXFYmidWD4+fkVcpBonSZ+fn5FOlsmTZpEfHw8QUFBjBkzhvj4eI6tPEb2nGzsYu0AWPjFQmxcbXCr7lYoYWpAUABNXJqQvCOZ5ORkjh49SsOGDR/3ZRE8Awzp87Bhw/T6VGV9jo6O1ksAWV76XFYt1rJu3Trs7e3p3Llzmcby8fHB3NwcAIVCgUKh4PLly6WyVfDsiIqL4mrqVQ5eO0hu+1yUzkq+nfUtdR01+77rOtbVRGlI8KrHq/xx7Q9OJ5xGISkY2Wwk3vbeZOdnY6wwZl3MOtJy0wxf6GE1GAtjC0L8Hm3LEVR+Cmp0UlLSC6fRFbGGLk+Ntra2Ztq0aXh5eWFkZETPnj3x9vaWndDnzp3j1q1bjB8/HmNjYzp16kS7du1KvFZlpbJsD9dSXvbo5vAA/ZwdC79YyKfrPmXhFwvxcfUhsm8kI5qOKKSjPq4+bBiwgbEtxzKp3SSD42qd4ZPaTaJvnb5YnbdCoVAwYsQI3N3dn2gOzxLhAClAmzZtSEtL4+zZs6hUKtavX4+jo6Pcrlar6dWrF02bNuXmzZvs3buXBQsWsHv3bgDq1KnDwYMHSU1NJSQkhKFDh3L79m29a/z000/s2rWLq1evcurUKVasWFEmG0uy4fz58yxevJhjx47x4MEDdu/ejZeXF5GRkXqe8IkTJxYaOzMzkxYtWgCa/V1TpkyhRYsWeolt1q5dy5EjR5gxYwZ+fn6sX78egLFjxzJ27NhSz+PChQsYGxtTr149+VjTpk1L5dEuCe0XI4VCUXJnQbGUl1hrHSlRUVGFHCRap0lUVJSes0U3+kQbiZWVlYVKpcLY2BgjhRHpieksW6AJMfVx9cHR0lEvrFo79uyg2QSoA+ji1wV3d/dKczdAUHoM6bPu1ouqrs///POP7AApT31+HC1OS0sjODiYefPmPdZYY8eOxcrKigYNGlCtWjV69OhR5LUElQNtZa0eL/Vg8PDBHI49TGBgYKFqAjsv7yQ+LZ74B/HkS/mk5KQwcc9EutXpxtsN3iZPnUeeVHTkh7lC4xxzV7rrbcsRVH4KanRycvILp9EVsYYuT40uSEJCAhcuXMDHR/M3ZijqXZIkYmJiih2nslJe28OfZOuK7raS8rJHu1VRm7S6T4M+RPaNJCouCkBPNw1F9OmOU7Cv7rha7PPsqX2hNjXNazJy5EhcXFyeyP5njXCAGEDrwd6zZw8NGjTQC087duwY9+7dIzg4GDMzM2rXrk1gYCDr1q0DYMCAAVSvXh0jIyMGDRrESy+9xNGjR/XG//TTT6levTqOjo706tWL6OjoMtlXkg3Gxsbk5ORw5swZ8vLy8PLyok6dOqUaOzMzk4ULF2Jra4uDgwN3795l165deo4Ef39/EhMTiYqKIioqikGDBgHw/fff8/3335d6Hunp6djZ2ekds7Oz48GDB6UewxDXr19n//79DB8+/InGEWgoL7EujSOlYJ+JEycSHx/PxIkTmTNnDg4ODlhYWODg4MCSJUuYM2cObtXd8OjpwbDNw9h2bhs3bt6Q9z5q0SZr2rt3r7zweNa5UgSPR0F91g2/rOr6HB0dzd27d8tdnx9Hi6dOncqoUaPkbWllHev777/nwYMHHDx4kH79+skRIYLKy5/b/mTfp/voldFLrq4FhqsJhHQMwcPGA8XDuOpsVTbh/4STkp0i5/jQtgGYYCI7PxpXa0wDpwbM7zb/6U9S8MToarSFhcULp9EVsYYuT43WJS8vjyFDhjB8+HC5fGmDBg1wdXXlm2++IS8vj//973/s37+fzMzMUr0GVZUn2bpScFtJeVFw3PK6TsFxrl+/TkREBJaWlowcOVIvMOB5RThADODv78+aNWtYsWJFodC9a9eucevWLezt7eXHrFmzSEhIACAiIoJmzZrJbTExMYX2/umGDFlZWZGenl4m+0qyoW7duixYsIBp06bh6urKu+++y61bt0ocNycnh+zsbE6dOkVaWhobN27k8OHDmJoWTlJWVlavXi0noOrevTugSVKVlqYfApuWloaNjc0TXSsiIoL27ds/dxmJqzolOVIMlbfVLhoUCgWBgYFYW1uTlZUlv28CAwMZvGww0R7RbDizgen7pyMpJbCFRu80IiwsjBo1avDBBx+QlZXFkSNHuHPnjoj+eI55kfX57NmzNGzYsNz1uaxaHB0dzW+//cb48eOfaCxjY2Pat29PfHw8S5YseYIZCJ4GRX0B0L2DqL17GBUXhae9J0YKI2zNbDFRmKBWq/n9yu/yeboOkHwpHxSQQw4x92LoVrcbfRr0eWpzE5QfuhpdcNtxVdZoSZI4e/Zshayhy1OjtajVavz9/TEzM2Px4sXycVNTU7Zu3cqOHTtwd3dn3rx5DBw4EA8Pjyeex/PMk0RDF9xWUl4UHLe8rqM7zpUrV4iMjMTGxoaRI0dib29fDpY/e4QDxACenp54e3vz66+/0q9fP722mjVr4u3tTUpKivx48OABv/76K9euXSMwMJDFixeTlJRESkoKjRo1KvckqsXZoOW9997jjz/+4Nq1aygUCrksaHFbQmJiYlAoFNSuXRuA/v37U6tWLTZt2vTENg8ZMkROQKXNNl2vXj3y8/O5ePGi3O/kyZNyGN7jEhERIaI/KpiKyKqtu7jWjv/WW2/h4eFBaKgmO3VwcDDGxsaoVCp5Ea4Nyx7QcAAhHUNwdHLEf4U/C79YyLRp07h16xb/+9//GD58ONOnT69Ue0EFZedF1mdzc3M5UqI89bmsWhwVFUVcXBy1atXC3d2duXPnsmnTJnx9fR9L1/Pz80UOkEpO7N1Y6verj5O7E/X71WfbuW2FSiNqw7xDD4Wy6vQqrqVcQyWpSM1NJV/KJyk7iYz8DBQoMMIINWrNiQ9zflibWONXy48BDQeInB/PMboaXfDLUlXW6KysLMzNzStkDV2eGg0aZ82oUaNISEhg06ZNhZw0TZo0Yf/+/SQlJbF7926uXLlC69atDV3qheFJoqGL24LyJBjavlIe19GOY5Zqxpo1a3B0dGTEiBHY2tqWh9mVgqfuANm1axf169enbt26zJ49u1B7amqqvDfPx8eH5cuXP20TAVi2bBm///471tbWesdbt26Nra0toaGhci6CmJgYjh07RkZGBgqFQt4XtXz58grZM1ecDaDZv/j777+Tk5ODhYUFlpaWmgy/gJubG1euXDE47j///IOlpaWewPfo0YOff/653OcAmkRM/fr1Izg4mIyMDA4dOsS2bdvw9/c32D8/P5/s7GxUKhUqlYrs7Gzy8/P1+vz555/cvHlTrv5S3LnPmspWmqssPG4oYHFz1vWu6+YL0T739/dnxowZDB48WO4XFhZGtxbd6JDYgYh+Edw9eJcHNx7QIbEDTmonXn31Vezs7Jg6dSrVqlWrNKWCKytCn5+citTnRo0a6R0rL30uqxZ/8MEHXL58mejoaKKjoxk9ejRvvfUWu3fvLnGsu3fvsm7dOtLT01GpVOzevZu1a9fSqVOnJ55HZaBgCcHStlXE9crzuuEnwjnpcZJa02px0uMkk3+bzPd/f887P70jl7vtvaY3i44s4kbKDbrX6Q4KMFGYFBpLgQJnK2dsjWyxkqywN7Wntl1tVvdfzb6R+/S21wieT7QabWSk/zWjKmt0ZmYmjRo1qpA1dHlqNMCYMWM4e/Ys27dvx9LSstD5p06dIjs7m8zMTObOncvt27cZMWLEE8+jsvAs198Vqdmxd2MZtnkYw7YMe+LPmbNnz7Ju3TpcXFwYPnw4SqXyicarbDxVB4hKpeKjjz5i586dnDlzhrVr13LmzBm9Pv/9739p2LAhJ0+eJCoqigkTJpCbm/s0zQQ0iZhatmxZ6LixsTHbt28nOjoab29vnJ2dCQgIIDU1lYYNGzJhwgReffVV3NzcOH36NO3atSt324qzATSh0pMnT8bZ2Rl3d3fu3r3LrFmzAE35wa+++gp7e3vmzp2rN250dHQhIezWrRt79uwplcNg9OjRjB49ukxz+f7778nKysLV1ZXBgwezZMkS2aPdvXt32W6Ar776CktLS2bPns2qVauwtLTkq6++0htv5cqV9OvXr1BYoKFzCybWetpUttJcZeFxQwENzVm3vK3WOWHIGbJ27Vri4+PZsWOHfK621vno0aOp3bE2H3z4Abm5uXwx9Qvaf9ie3b/tJiQkhAkTJpTPxKswQp/Lh4rU5yZNmugdK099Lk6LQV+PrayscHd3lx9KpRILCwv5i0txYykUCpYsWYKHhwcODg58/vnnLFiwgD59qsZ2h+L2YFfEPvDSjFke19XN7TG08VDcle7kqfK4nHyZ8BPhvPfv97gScgX1cTUJmQmk5KRwPfW6wbHUqLmbeZc0VRqZikxS8lOIfxDP9P3Ty9U5VBV5XpzUL6JGZ2VlPRcafe3aNX788Ueio6PlNqVSyerVq+WxIiMjqVatGq6uruzdu5c9e/ZUqTxNz3L9XZGaHX4inA1nNrDhzIYn0vvTp0+zYcMGqlevzvDhw/Wqz1UZpKfIn3/+Kb3xxhvy81mzZkmzZs3S6zNr1ixpzJgxklqtlq5cuSLVqVNHUqlUxY7bokWLMtkREhJSpv779u0rU//nGTHX8qE077GlS5dKHh4e0tKlSyvMDi3P4vcakxAjdf60s+RW3U2eo6E5e3h4SIDk4eEhn+e/yV/y3+wvxSTESEuXLpXcqrtJDV5vINnY2UgKhUICJCMjI6lGuxoSCiRA/n/u3LlS7d61JYWtQm/c55GyatuTUFH6LEllm0dZ9VmShG5VVSpqro/zHistMQkxUtDOICkmIaZMbSXNtajPi+LGLEuf0lxLF/9N/pLpDFPJYbaDtOjwIklhp9FbbJF6rekl1V5QW2Iaksl0E4lp6D3mrpkrEaL5WTFNIZnMMJEcvnaQ7GfbS0E7g0plY2XhaWp0fn6+VLt2beny5ctSTk6O1KRJEyk2Nlavz3/+8x9p4sSJkiRJ0t27dyUHBwcpJyen2HErcg0t9KpqUtFzrUiNlqSyrb/Le64Vodm65+mun8vKvn37pBMnTkjTpk2Tli9fLmVnZ5d5jMpEcdr2VCNAbt68qZeN2MPDg5s3b+r1+fjjjzl79izVq1encePGLFy4sFAInUBQFajq2zHCT4Szb+U+Em4lyF52Q3MeFTQKpbOSZm2aUbNmTcb9Zxwbzmxgfcx6Xl/5Oq4dXBm8bDCJ3RORTCV5P7Barebm8ZvQGFCAUx0nLBwtsLG3YZDfIOo0r4ORkRF+fn7PYPbPH0KfBYInpyzlBstCaRKQlsWm4kLAS7o7qo3S8LD1ICMvgyV/L8HsdTMUdgqcujkRFReFhIStmS0KhUIv2anMw0M+zj583Opjlr+9nBFNR4jcH8Vw9OhR6tatS+3atTEzM+Pdd99l27Zten0UCgUPHjxAkiTS09NxdHTExKTwNiSB4EXmWay/tZr757Y/H0uzi6LgdhknKycmtZskn1va7TSxd2O5mnSVsG1h1K5dmyFDhlSpqJ+CPFVVlAwkMiqYUGj37t00a9aM33//ncuXL9O1a1c6dOhQKPHK0qVLWbp0KYCcK6C01K9fv0z909PTy9T/eUbMtXwo63usoklPT2fjxo3cvn2batWq4ezsXOHXfNPkTV7+7mWyk7OxsbHhu+++M3jtui/V5avFX6G+o8avjR+mpqYMrTWUlOwUVJKKWzG3aK1qTS2zWtiE2pCRnqHJ/fIwcR5GgEqTudzFxQULCwvqetelmkk1cnvlYmZmRlRUFImJiU91/s8b5anP8Pga/Th/O0K3qiYVNdfKps9Q8lznz58v61d52J6dnU1QUBDZ2dnyeNn52SRmJjJn7hwS7yZSrVo1tu3axq30WzhZOpGrysXZypnEzERa5LXgleqvkJOfg625LRYjLcj2zyYtJw0FikcJTg3gYe7B3HpzUaDA0dIRLwsvuAN9LPpw78w9os48+fyqIoac1EeOHNHr8/HHH9O7d2+qV6/OgwcPWL9+vXBSCwRPQGJiIjVr1tSrUKjFUPXCotB1LOv2LWoM3eNt+7Ql/EQ4Ab4B+Lj6sO3cNqbvn87IZiNZHr2cC0kXWH16NW1qtOGvm38Bmm2L7254l5jEGMyMzEjKTNKsmSWY1H4Sl+5fYvr+6YR0DKFPgz7M3DGTVyxeIc4hjkWDF1V5x+lTnZ2Hhwc3btyQn8fHx1O9enW9PsuXL2fy5MkoFArq1q2Lt7c3586dK5R9+IMPPuCDDz4AoGXLlmW6yztt2jQGDx5c6v5RUVEvzF1kMdfyoazvsYomKiqKMWPGEB8fX+jvsLQUJ/SG2sLCwpg5YybBwcEEBwcXee3Yu7GEnwjn0sZL7Nyyk8GDBxP5aaRcZQAFpH2dRlJCElZOVvQP60/tK7X54dsfaPROIyxvWfLr5l9p1KgRY8aMwc3NjU6dOnH58mXmzJlDcHAwfn5+1KxZ84nmX9UpT32Gx9fox/nbEbpVNamouVY2fYan/3u9ePGinj4CjN81nlXnVzG08VDmD5oPQK2htbix/QbKrkosXrHA29abU/dO0cSlCf7N/Jl/aD5p2Wl81Poj1sesJy41rljnB8DcenP5Jv4b3qj7Bv0b9Gdb3DZ5YS8omufxJqJw2FZNKnqulclJrVarCzmLtRhyJBdFUU7srKwsgoKCyMrK0juuHftB+gP+t/Z/1LKpxT+5/3DP9h63Em8xxHgIuX/l4m/jj+QtISFhrDDmTa838cCDfw7/wwjHEeCoGc9GsiE9Nx0kOPbnMVJzUhliO4RbMbf45dIvvGHxBkoLJR/U/4A//vijXF67ysxTdYC0atWKixcvcvXqVWrUqMG6detYs2aNXp9atWqxd+9eOnToQEJCAufPn5dLSgkEgsdHm1D0cUvATpw4kZSUFCZOnFjIAWLIs617rLhra0P9agbWRKVSyR8AfRr0Yfr+6UTficajiwc5u3PIbpvNhjMbGN1iNHdu3gHA3d0dtVrNjRs3GD58uHxHLDAwUM/O4OBgJk6cSEZGBmFhYVV269HjIvRZIHhx0NVHrRPaz8sPQG8bSuZvmZAGJgdNGBowlP8e+y956jxO3T3FleQrxD+IB+CbP79BLamLdH4YY0w1q2r4N/fHJMmEzPxMnCw122VWnV4FwPxu8ytwxs8/z+NNROGwrZpU9Fwrk5N648aNLFiwQM9ZrMWQI9kQujcJ33nnHb22r7Z8xb6V+3h9+Ot8+umnemP/e8a/SU9Ph2xQOis5HHsYH1cfUs+lMuDVAeSl5KF0VjLrl1nMPDCTlOwUFAoFPi4+hHQMIXRvKDGJmkpK3nbeNHJrRExCDCjgbsZdrEyt+MTzExL+TuC483G6uHQhzjKO+X5VX4ufalyciYkJixcv5s033+Tll19m4MCB+Pj48MMPP/DDDz8AMHXqVP78808aN25M586dCQ0NFeHqAkE5UNKex5LKgmnvNBW84wSGq8LoHit4be21ps2bJu9NDA4OxsHBgfT0dMLCwth2bhu3Um9hrDDGvKU5Ll+6UNu+NtI8icjlkfRe25vRE0eTnJqMiaUJYyeNNVjOTXf+SqWS5OTk57LyTkUj9FkgeHHQ7gvfdm4b3T/rzsJ3F/Ldku9kJ0SXcV2wtbclMy0TBwcHxk0cB0B9x/qaARQQfTsaNys3AEyNTDFSGKE0VWJnalco74cKFc09mpOVn4UkSdR1rEuAb4BcXUbk/igZXSd1bm4u69ato3fv3np9tE5qQDipBYJywNnZuci1c2lziRSXV2nhFwv5dN2nLPxiYaGxD8cepvMHnXGr7sa3s76Vo+T6NOjDf+f8Fw8PD76d9S2fvPIJ+4bv491G71LbvjZXU64SFRfF6Y9O06teL4wVxjRya0QdhzrcTr/NzQc3qedUj5l1Z6I6o6JFixZ8PeBrnCydXhgtfuobfHr06EGPHj30jumWfapevTr/+9//nrZZVZbr16/TsGFDUlNT5TrmAoEhitqfqCU0NLTIKI6C0RZFHSt4ra9DvibfJJ/Tw0/z28Lf5OMTJ04kgwzyauXBBbjEJegKpgdNyUvJI+enHLYf2g63AAmM7IzIaJpR4hyfNAqmqiP0+eki9FnwrAg/Ec6KkytYfXo1ib8kIqVKRP8Uzfju40nKTGLfyn2oUzXRHE5OTqT6pLLi5AocLRwxNzInR5VD1PUorE2ssTS2JEuVBUB6XnqR19x1cRfz3pyHTbYNjcwbcen+JaLiosT2l1Ki66RWqVS8//77spMaNFo9depURowYQePGjZEkSTipnxCh0YLyoDRR0IbwcfXht4W/wcLCbQXX2D6uPkT0jZAj+rSOjK87f00dhzry86TMJADaqNtw++RtXnnlFbp164ZCoSDRNvGF0WKRGekhXl5emJmZkZiYqHe8WbNmHD9+nLi4uKdmy+rVq+W63JaWlhgZGcnPlUplmcaqVasW6enpjyXcUVFR8rVtbGyoX79+oZryCoUCNzc3TVLKh+Tn5+Pq6qoXKRAbG8sbb7yBg4MD9vb2tGjRgl9//bXQdXQff/31V6nsvH//Pn379sXa2hpPT89CYfsFmT9/Pu7u7kRHR/P++++Tk5Mjt/n5+WFhYSHbUL9+fbnt8OHDdO3aFUdHR1xcXBgwYAC3b98ulY0FKSna4mmPA4ajOHQpr6zZYWFhZGRkYG9vj5HaCHWqmmMrjwGPKsKoJTV5KXlwCsjWPIwPGlO9p0647000iVABz6aepfJaV/XKO1WV4vRZoVCQm5v71GwR+vx09NnOzo5r167p6bOWixcvYmFhwdChQ+VjZ86coWXLljg4OODg4ECXLl04c+ZMqWysrOhm7y8pk39p+uoer+1Qm+y8bJIPJaPIVWBrZ0uzgc0021EU0Hpwa0ytTbG1syU4OJgA3wC87b2JS4kjR52DEUZwHDLmZJB1NKvYeRijeX/nSXksObYEEyMTdl7ayfT901l1ehXhJ8LL5wV7AejRowcXLlzg8uXLfPHFF4DG8aF1VGud1KdPnyYmJkbvb6QiKUqjz5w5g0KhEGvoKqbRp06dws7OrtAaWoshjdZl+vTpKBQKfvvtt1LZ+LxRcG2uu/Z80nW7bvSers5vO7eNhosbMmXvFPy8/Ag/Ec53R77Df4s/fl5++Lj64OPqw4q3V9DfuD93Y+7Srl072fnxoiEcIDp4e3uzdu1a+fnp06fJyir+g70iGDJkCOnp6aSnp7Nz506qV68uP09P17+7olKpKtQW7bXT0tKYP38+gYGBnD9/Xq+Pvb09O3fulJ//+uuvODg46PXp1asXXbt2JSEhgbt377Jo0SK9pFwF55iens6rr75aKhs/+ugjzMzMSEhIYPXq1YwZM4bYWMOLxN27dzN79mz27t1L48aNuXLlCiEhIXp9Fi9eLNugO9fk5GQ++OAD4uLiuHbtGjY2NowcObJUNhakpDKDT3sceHrOgRkzZpCcnIxSqcTC3AKAB2kPCAsL44/rf5CZl0l13+qFFhwvNX+J66nXH6mWPXIpxXtn7r0wXusXFaHPhanK+hwXF0dOTk4hfdaO2apVq0KvxcaNG7l//z6JiYn07t2bd999t1Q2VlbCT4TLDgLdnx+3r+7x5dHLyVHloI5So85SY2tjy8IvFjK08VAmtZvES2+8hPEUY/os70PbPm0JPRSKh42HnONDjRr2A2lo/i8CY4ypZV8LI4VGuOs61cXZypmhjYcS0jFEbH+pQhjSaEOJWysaodEVr9H16tUjLi7O4BpaO2ZBjdZy+fJlNm7cSLVq1Upl3/NIcWtzbdvUaVNlB0ZZnCJaHdc6kEMPhTJs8zBGbh3J2aSz/HrxV7lt5oGZRN+JJnB7ILF3Y1GpVGzZsoV//vkHPz8/unTp8kI6P0A4QPTw9/cnIiJCfr5y5UqGDRum1ycnJ4fPP/+cWrVq4ebmxujRo+VFeHJyMj179sTFxQUHBwd69uxJfHy8fK6fnx9Tp06lXbt22NjY8MYbbxTylpfEiBEjGDNmDD169MDa2pp9+/axY8cOmjdvjq2trSavwrRpcv+4uDgUCoXsXX5cGxQKBT169MDR0ZFTp04V+7pFRETovW6JiYlcvXqVwMBAzMzMMDMzo127drRv375MczdERkYGmzZtYubMmSiVStq3b0/v3r2JjIw02H/lypWMGjUKHx8fjI2NmTp1KitWrCjVtbp3786AAQOwtbXFysqKjz/+mEOHDpV4niFhKynaorSU1ziPQ1kEW7evn58fxsbG1KhRA9C8tyRJYsaMGUSvj0adqibuZBwfBH+A7jby66euY3LABG2OPStjK5q83wQrJysmTJ5QEVMUVCKEPhdNVdRnBwcHqlWrVkif161bh729PZ07d9Y7bm9vj5eXl6wnxsbGXLp06Ynn8KyIvRtLUlYS3et0L1WujADfAJrGN2XtqLXYxdoZ7Ks7RkjHEJq5N+PDzz6UP0N8XH0I8A0g/EQ4N1JukJOfw08xP9F0SVMiT0Wy4+IO7MzsHg3YEbB9+H8ROFo5Mv/N+QxpPAT/xv583flrLEwsmN9tPn0a9GF+t/nCeV1FMKTRTk5Oen2ERlcNjbawsMDBwcHgGroojdby8ccfExoaipmZ2RPbX1kpbm2ubWv0TiPZIV2Wm5laHdc6kJFgw5kNpOel42DuQI+XeshtU1+birOVM9n52YQdD2Pjxo2cPn2aLl264Ofn98I6P0A4QPRo06YNaWlpnD17FpVKxfr16wuFb02aNIkLFy4QHR3NpUuXuHnzpvyGVavVjBw5kmvXrnH9+nUsLS35+OOP9c5fs2YNy5cv5+7du+Tm5jJ37twy27lmzRq++OILHjx4QPv27bG2tiYiIoKUlBR27NjBkiVL2Lp1a7Hnl9UGtVrNzz//TGJiInXr1tVre/vttzlw4AApKSmkpKRw8OBB+vTpI7c7OTlRt25dhg4dytatW0lISCjTfGfPnk3Pnj0Ntl24cAFjY2Pq1asnH2vatGmR3uvY2FiaNm2q1zchIYGkpCT52JQpU3B2dqZdu3bFlrU6cOAAPj4lL9wMCVt5RVuUxzjFhVYXbJs2bxo2Ljb0GtBLLqtbnGD7+/tjYmJCUFAQ8fHxfDTpI3bu2YlKpeLYsWOkpKRgZWOF0llJr8BeWHWxAlvIb5/PYbfDWPezxtjMGBSQ65FLjQ41MDMzQ6FQ0LhuY+7vvk+/7v1YtmBZuWwDElRehD4XTVXVZ0tLSz19TktLIzg4mHnz5hVpj729PRYWFnzyySf8+9//LtNcygNDjuHHCXkOPxHOzks7cbJykkOXi3MW+Lj6cH7zeRJuJbBswTKDfXXH6NOgD5F9IzFvbc6u47to26ct43eNJ/RQKKtOr+Ja2jUkJHLUOajQ3CVXo0al1rlj3gL47OH/BlCgwM3ajU1nNjGp3SQi+kUIZ0cVxpBGOzo66vURGl21NLrgGrokjd6wYQNmZmaFco1VNYpbm2vbtBF3Ab4BZbqZqdVxrQN5UvtJDGg4gHcbvcvB9w/y8+Cf5TZtctSRTUdS7XY1zp49S/fu3cvFefa8IxwgBdB6Yvfs2UODBg3ku9SgqcEeFhbG/PnzcXR0xMbGhn//+9+sW7cO0IhU//79sbKywsbGhi+++IL9+/VjQ0eOHEm9evWwtLRk4MCBREdHl9nGPn360K5dO4yMjLCwsMDPz4/GjRtjZGREkyZNGDx4cKHrPq4Nt27dwt7eHktLS/r27cu3335L8+bN9fpYWFjQq1cv1q9fL2clt7CwkNsVCgX79u3Dy8uLCRMmUK1aNV577TUuXrxY6Dq6j4wMTVLLyZMn88svvxi0Lz09HTs7O71jdnZ2PHjwoFT9tT9r+4eGhnLlyhVu3rzJBx98QK9evbh8+XKhcU6dOsWMGTP45ptvinzttDzLKI3SUFxodcG2ebPnkZ6Yzo7NO1CpVBgZGVG/X305hM/R0REHBwd5ob9mzRpUKhWZmZmYOpiS1z4Pqy5WeHh4UK9jPYzsjHDs5Yiqg4ofv/2RhIwEHKY44JbkRvSYaLK3Z2Ntaa1JdHrViH8N/xeurq5IksTff/9NfHw8a9euLbdtQILKjdBnfaq6Pmu3wWn7T506lVGjRlGzZs0iX5OUlBRSU1NZvHhxodfiaWDI4f04WxV1IzpK6zgp62eNrr6H/hHK939/zx/X/qB73e40cmmkyfNRgPT8opOcFkRCIvZeLBvObBB5Pl4QCmq07l1+odFVT6MLrqGL0+j09HT+/e9/s2DBgqJerhcKXYd0YGCgnCi1NHq/7dw2fH/05bsj3zHuP+PYPmY7UZuiuHS/cNTjS/Yv4XbLja03ttLQryGvvPJKkXlEXiSEA6QA/v7+rFmzhhUrVhQKr7537x6ZmZm0aNFCFphu3bpx7949ADIzM/nwww/x9PTE1taW1157jZSUFL09hu7u7vLPVlZWhfYjloaCwnLkyBFef/11XFxcsLOz44cffig2JK8sNlSvXp2UlBTS0tL49NNP+f333w32GzZsGBEREYVC97R4eHiwePFiLl++zLVr17C2ttbrp72O7sPa2rpIu7QolUrS0tL0jqWlpWFjY1Oq/tqftf1feeUVbGxsMDc3Z/jw4bRr105ONKXl0qVLdO/enYULF9KhQ4cSbayo3BrllQC1uNBq3bbYu7E0eqcRVk5WvNXvLTw8PHj949c56XFSDuFLTk4mJSVFXujX61gPFODQ2oHPf/oclw4u/OuTf3Hjxg2avdIMSZJIyU4hd18u6lQ1udtySfkzhZt/3kSSJFS5KtJS0zCzMGPAewPYfHAzaQ/SsLS0xMzMDAcHB1q2bImxsXGF1qQXVA6EPutT1fVZrdbsd7OxsSE6OprffvuN8ePHl3hda2trRo8ezbBhw7h7926J/cuTkkqClxbdiI5JkyY9ltaXlAzVz8tPdrJc3HMRlVrFtdRrAHzd5WucrDTbFxQoMDUyhePAt2j+LwZThSlOFk54KD3o6NmRAQ0HiDwfLwhCo/Wp6hqtu4YuSaNDQkLw9/fH29u7RLuqAmVdo5fkKNfV8+n7pxN9J5p//e9f7F2xl5S7KdzYfoPp+6frnZOTk8OqVavYfnM758zO8Ve2JjFuwTwiL6KDWjhACuDp6Ym3tze//vor/fr102tzdnbG0tKS2NhYWWBSU1Nl8Zs3bx7nz5/nyJEjpKWlceDAAYByTwJVcM/We++9R+/evblx4wapqamMHj263K9pbm5OaGgop0+fNhga2KFDB27fvk1CQkKJoVU1a9bko48+IiYm5ontqlevHvn5+Xqe8JMnTxa5NcXHx4eTJ0/q9XVzcyu0T1WLdj+5lmvXrtGlSxemTp2Kv7//E9v/JJRXAtTiQqt128JPhHOp9iU+WPUB2zdsNxjCZ2VlhUKhkJ0RG9dtpPmS5kj9JHZd3oWExJXkKwBs+mETUprEg80PsK5rrcn3IYG0X4Lq+nYYWxrj0seF38N+Jy01jezsbLKysrC2tubmzZuoVKpitysJqgZCnw1TVfU5MzNT1ueoqCji4uKoVasW7u7uzJ07l02bNuHr62twLLVaTWZmJjdv3nzieZQFQw7vx3WCax0nkiSVSusLfiaUlAw1Ki5KdrJc/fkq9RzrYWJkApJG+8N6heFi5aI5BqVKegqaai+17GtxY8IN9o3YJ7a+vEAIjTZMVdVo3TV0SRq9d+9eFi1ahLu7O+7u7ty4cYOBAwcSGhr6xPOojJRmja7rJCnJUa7V7dBDoXjYemBvYU++Ol/OxaTsqmRks5FyZMeQjUPouLAjJ26cYHLXyYxoPkJ2RBfMI/IiOqiFA8QAy5Yt4/fffy/kPTUyMiIwMJDx48fLd5Vu3rzJ7t27AU0ImKWlJfb29ty/f5/p06cXGrsiePDgAY6OjlhYWHD06NESS1g9LmZmZkyYMMHgH7NCoWD79u38/PPPhT5ckpOTCQkJ4dKlS6jVahITE/m///s/2rRp88Q2WVtb069fP4KDg8nIyODQoUNs27atSOfEsGHDWLZsGWfOnEGlUvHVV18xYsQIQBM6vXv3brKzs8nPz2f16tUcOHCAN998E9D8rjt16sRHH30kl5x7FmgF08/P76lurTEUKfLntj/ZGLiRP7f9CWi8zZIkyc4IH1cfIvtGMqLpCD2hDQsLw0z9MDRWAtUFFSHfhGDlZEW9vvVQ3H/4HjIBU3tTJn4xkQDfAM1dSDQLInt7e4KDg+WkqiIC5MVA6LNhqpo+Jycnc/v2bVmfP/jgAy5fvkx0dDTR0dGMHj2at956S/797tmzh3/++QeVSkVaWhqfffYZDg4OvPzyy088j4pEe1cvOz+7UFvbPm15J+wdxn05DrfqbvJ2w6IiOwouoAtG8GnP0T2uPWf0Z6NpWb0lgxoNolWNVvj+6MuR+CM4WznTzqMd5kbmpUp6CmBpbElKdgrbzm3Tm+OLGGb9IiI02jBVSaOzs7NJTk7WW0OXpNF79+4lJiZGbq9evTo//vgjH3300RPPo6IpKbeTIY0LDg7GrbobypeUuNdwNxgJMnHiROLj45k4cSKuHVxx+dIF1w6u8haXbee2EXs3lt5re7MxdiM2ZjbcSL3Bnit7aOrWlHpO9fDr54ffAj9qvl6TPZf3sOLkCgK2BbA+dj1Hso5ww/MGfdrqJ5wumEfkRXRQCweIAerUqUPLli0NtoWGhlK3bl3atGmDra0tXbp0kUtaBQUFkZWVhbOzM23atKFbt25Pxd7vv/+e4OBgbGxsmDFjBgMHDqywa73//vtcv36d7du3F2rz8fEx6DU2MzMjLi6OLl26YGtrS6NGjTA3N9fLHH3r1q1CNcw3bdoEwKxZs+jevXuRNn3//fdkZWXh6urK4MGDWbJkiWzH9evXUSqVXL9+HYBu3boxceJEXn/9dU6dOoWnp6f8IZuXl8eXX36Ji4sLzs7OfPfdd2zdupX69esDEB4ezpUrV5g+ffpj15QvidKEzGm9ylFRUeW2taY01y3o7ABNQrP4+HgmTZrEjBkzUKlUsjNCO15Bof1z25+MGTOGjLQMeRxTI1OmTZhGRmIGseGxWCosAbCxtCE3OZcatjXo1qIbA/sOlPMCKJVKAgMDiYqKkiNAwsLCOH36tEiIWoUR+lw0VUmfPT09MTMzk/XZyspKvnPo7u6uKaNtYYGLiwugcWAPHjwYOzs76tSpw6VLl9i1a5feXvrKiPauXmJm4ZB7bVuqTyqDlw2WtxsWFdmhG2kSezeW0D9CScpK4tL9S/hv8WfFyRWEHgol9FAo0bej6bW2F9uV29l1fBepPqlsPb+VX87/wpTfpvDPnX/4+tDXnE08S9T1KDLyM0pMegqa7TJGCiPiUuLkcOySSvgKqhZCo4umqmj0+fPn8fT01FtDl6TRTk5Oeu3GxsY4ODiUuI4ur+3eT0JJuZ0MaVzbPm2pHlyd8yc0UXYzZswo5CjROrvUkprA7YFE34lm+v7p8haX6funE3oolF8u/EJ8ejzXUq5xeudpsudkc3rnae5l3aOZezMSMhI4m3iWmHsx1LKpxYPsB5hiipmRGXZ2dsIJbQipCtCiRYsy9Q8JCSlT/3379pWp//OMmGv5UNb3mBYPDw8JkDw8PIrss3TpUsnDw0NaunRpqcfdsGFDoXN0xynNdQ31sbe3lwDJ3t6+1ONp27QPhUIhLV26VFq6dKlkb28vWVpaSnUb15UsHCykNqPbSDEJMfI5SmelFDI3RLKxs5HMleZSyNwQaejQoZKxsbE0dOhQycPDQ5o7d26x83ieKKu2VVbKMo/H+dsRulU1qai5Pq4+VwQxCTFS0M4gaeeenUW2xSTESFvPbpWa/9BcWnR4keS/yV/y3+wvbT27VW6XJEnuoz1uMdNCsvjKQnr5u5clsxlm0svfvSz5b/KXLL6ykJiGxDQko2lG8rgWMx8df9KHxUwLadHhRZIkSVLI3BBZuyWpar2Hq4JGV+Qauir9rktCzLX80H2PlWZ9WpHs27fP4Lpb95iuVmsJ2hkk2c+2l2oOqSm5VXeTli5dKgXtDJKc5zhLQTuD9Mbo/GlnyWaWjeQyx0UaM32M5OTmJLm/6y75LfeTai+ordHVXkhGdkaSqZWpBEimDqaSyxwXadHhRZLfcj+JaUjVvqkmBXwTILWd3lb6MepH2aaC1y1urlWJ4rRNRIAIBJWI0iTLe5z95Ldv3y7We12a6xrqM2fOHGztbMnMy+Rm2k0W71nMEvUSegX2wq26Gx49Pei9tje91/Zm2OZhxN6NZVTQKJTOSswsNFtgbGxt+P3K73zw4QekpKSQlZXFrZu3YAL8fetv2vi0oUaNGiiMFOTWzCXVJxXJVCInPYd5s+fpRYAEBwdjZmZWaSvuCAQCgS7aCDkLk8KRKro5mL5b8h0nJ53km+++YeflnThZOhEVFyXfdQwLC2PAqwP4Z8c/TN8/nQDfAAY0HMCAhgOo61QXIyMjWtZoSe2rtTGab0S9K/XwtvfGy86LqylXWR69HGMjY/naxhhrtr0UpITUCAoU1LavjYWphZzvadmCZaQnprNswbIneq0EAsGLx9OqpFhcpElJuZ0M5dIL8A1gRNMR7Px2J3du3qFtn7YkZSbRvW537GLtcK/hztervuZ+1n3qOdXj7fpv061uNzb/uJmkhCTS9qRx8NpBrqVe01Tk2g/qVDX5Uj4WjhbYdbXjXuY95h+eT0JGAkYKIxIyEtiRsYPpb0/ng44fyDZptz36efmJSJCHCAeIQFCOaAX0+PES0uQXQUVVjKlWrVqxlQkKXtfQB4FuH207gNpUTW5GLvNmz5PD9tbtWMfdO3c58ucRtl/YzvYL21kbs5Z3fnqHCJMIMtpmkEsuWECaZxrrQtfJC2sTSxOGjxvOgIYDMDtkRnpiOn///TeSWsLshhkBvgFMmDwBpbOSCZMnFJpH48aNy/31EwgEgvLiccKRYzbGoE5Vk7k3k6GNh2IXa8faUWtpGt+UAN8Apk6bSl5KHiYHTQjpGIKPqw8R/SKI6BvBqOajqKasxqHrh1j8zWIykzK5uPYicZ/F4fG7BzanbbjwxQWyDmcBGueHpYkluepcfSMkNMmqdTDCCGsTa8yNzVGaKlnYbSE/D/6ZEU0fJdyr7KXgBQJB5aWi1sUFKa/CAloKOkXCT4TLzutlC5Zpkk8fvEpmUiaR30XiZOXE5tWbSU5NxsHBgSYDmmBibIKtuS2WppYouyoxtjNG6iLhPd2bV99+FWOFMZm5mZxLPIeJZIIttmSYZrDj1g6Dtug6zV90TJ61AQJBVUIroAXLij0rwsLCmDFjBvPnz+fGjRt6bYGBgXoOjxkzZjAqaBSpPqmsnbZW3rPYtk9bQv8IBQVMajcJH1cfeZ6TJk3CSG2EmbUZEyZPoHnH5honyN/RSGoJTgMPE8GrJTWXki+hOqZC+kXSLKZtgQvIzg9FLwUmrU0wb2lO8xvN2a7ajrmDOW+99ZYc4eHj6oPPBB+mTZimNxeBQPBiotUvrRO0sqPdLw4wv9v8Up0zc9rMR3PsFkjNwJok3ErAdLMpPgt9aPROI+6tvEfHYR3p06APoHG0hJ8IJykrieup11FJKmp2qUnKjhRUKZrSogd+OQA2PKru0gJUqEjPL1Das4DzQ4GCltVa8iD3AbO7zJavqUV3XrqfNQKBQFAZCQ4OljW2ItCtwNIwuCFTp03Fqq4VCbEJTJg8gQG+AwjfH05uRi6uDq6ET9fke/Lz8iMqLoqAgACm7J3Crxd/pa5TXb7u/DV1HOqw+cxmJCTyyKPLS13wcPQosqqLrg0vOsIBIhCUI1oBbdas2bM2BXjkkLl9+3ap+s2bPQ+LiRY0facppptNCQ4OJvxEOBvObAAFOFk6Mb/bfPz8/Fi7dq1cjtbDw4NpE6YRFhbGva/u0bpVa/7++2+69+2Oop6CP679QcaRDPJ25yFlab0daBbeD6tUWiotsTprRdL2JPb67mXj3Y2kpKTg4eFBZGRkhb5OAoHg+UX3zt3z8EX7cRahBZ0I2s+aUUGjGL9rPJ+M+YTG3Rvrjal1tHSv050eL/XgYtJFxnYby5c+X5K2Lk3joG4MeKJxfhRV3aWIyI/03HQSsxKJiosq5AARCASC54mKdtRqozAAfAJ9DF7r21nfyk4Y3f5afdU6PQJ8A/Bx9WFi04kknUhiF7vw9fJlWpdpxVZ00R3zRUdsgREIyhFtqF6LFsWkyS+G4vYg6rYV9XNBtKHH1apVK/a62n5+fn48mPGAoyuPyndTdfeSaxfX2rwb5ubm8nmOjo58+OGHxMfHc/PmTfLz89m+YTs/D/6Zg+8fxOwPM9n5oVAomPzVZLj1yIZWr7ci6VISAKdPnCY+Ph5LS0vZG1/WTOCVIXO4QCCoeJ63LRaG9otrMbQ9xtCx7CbZ5Hyaw6/KX/n+7++Z/NtkAnwD+O8P/8XE3gSHgQ78dvk3VCoVHrYe2FvY85LTS3x/7Hvy1HmYv2OO3X/sNBF6xVV3MeD8AE2UyP2s+zhbOuPn5feEr4hAIBA8P5R2famr3SWds+3cNr7N+Zbmc5rTtk9bvVK42rFCD4WSlKlZJ8fHx7Ny5UqaWzbn3Cfn2DV81wtZzvZxEQ6Qh3h5eWFmZkZion4pumbNmnH8+HHi4uKemi2rV6+Wy1hZWlpiZGT0RGVX4+LiUCgU5OfnF9ln2rRpmJqa8s8//2Bvb0/btm3566+/5PaoqCgUCgX9+vXTO+/kyZMoFAr8/PzkP+6xY8fSrFkzbG1tcXZ2pnPnzvLrp72O7nzs7e1LPZfo6GhatGiBlZUVLVq0IDo6usi+P/30E23btsXKygo/Pz+9tgsXLnD58mVcXFxwdHTkzTfflEuxAeTk5DB+/HiqV6+Og4MDY8eOJS8vr9R2Pi7F7UHUbSvq54JoHTLOzs6F2nTFODAwkODgYHZs2kFOeg4PUh/I4/m4+tAhqQP7Pt0nl8ANDg7G3t4ehUJBcHAwUVFRJCcnI0kaB4fyJSXbzm3TW7Q3eqeRfG0LCwvmh86H6g8PGMOB7QcK2Zibmyt7ycu6P7O893MKnh3F6bNCoSA3N7eIM8ufZ6nPSqWS6Ojox9JnLdu2bav0+qwd6+zZs1hZWdHQryFDI4fqOQDmz5+Pu7s7dnZ2/PXXX1y6dKlC7t4Vla9De7ygzj0phsophp8IZ8XJFbzz0zsM26JJJj3zwEzuZd7j+J3j5KvyuZJyhdBDofz47Y+oUlWk7E4hZncMyV8nM3vRbFadWsUvF37hfNJ58lR55KhzSM1NLd6YIpwfWu5l3iMuJY6ouKhymbvg+aUojT5z5gwKheKFWEOfOHFC1swXRaOLGmv06NF6Npqbm2NjY1NqOys7pV1f6up5SedM3z+dc0nn+PXir4SfCNcrhasda8OZDWw4s4H5++cTERGBlZUVI0eOxNHRsdznWNURDhAdvL29Wbt2rfz89OnTZGVlPXU7hgwZQnp6Ounp6ezcuZPq1avLz9PT00se4DEZNGgQzZs3JzExkddff50BAwbotbu4uPDnn3+SlJQkH1u5ciX16tUDHgnCDz/8wLx580hNTeXq1auMHTsWIyMjvevoziclJaVU9uXm5tKnTx+GDh1KcnIyw4cPp0+fPkV++XF0dCQoKIjJkycXaktJScHOzo7z58+TkJBA69at6dPnUQjv7Nmz+fvvv4mJieHChQucOHGCr776qlR2PgnF3cnUbSvq5+Io6H0uKMYzZsyQHRhax4YWbd8xY8bIDhOlUklycrLBPZMXTlxgVMgoFgxaQMdPOvLxjo857H4YeoG5gznZUjY5aTko7itQGClAZdjmli1bEhYWhoODA/fv38fe3r7Ud3mft7vCguIR+qzRzWbNmj2WPgNcunSJYcOGVXp91o7l5OREcnIyrt1cWXtmLT/+/SMAu3fvZvbs2ezdu5e4uDiuXLlCSEhIqewsK4YcErrHp++fXq5J5Qxl6w/wDcDb3psrKVfYcGYD4SfCCWgegDHGmBqbUt+5PoMaDgIJzDqZobBVYP+mPUYHjCANpP0Suv/ypQJf5I4D3z78X/vz3xTp/FCgwERhgqe9JwMaDsAu1k5E2wkMarR2TfE0eVYa7eDgQHp6+mOvoeH50WhJkood64cfftCzcfDgwYVej+eZ0q4vtXoe4BtQ4jkhHUNo4NSAHi/1IMA3gJCOITRzb0ZIxxB5rAENB9DNsxtW562wtbVl5MiRZXKACR4hHCA6+Pv7ExERIT9fuXIlw4YN0+uTk5PD559/Tq1atXBzc2P06NHyIjw5OZmePXvi4uKCg4MDPXv2JD4+Xj7Xz8+PqVOn0q5dO2xsbHjjjTcKectL4tatW/Tv3x8XFxe8vb1ZtGiR3Hb06FFatmyJra0tbm5ufPbZZwC89tprANjb26NUKvW80oYwMTFhyJAh3Lx5k3v37snHzczMePvtt1m3bh0AKpWKn376iSFDhgAaQXBycqJGjRp07twZhUKBjY0N/fv3p1atWmWapyGioqLIz88nKCgIc3NzPv30UyRJ4vfffzfYv0uXLgwcOJDq1asXamvdujXOzs44OjpiamrK+PHjOX/+vPzBtH37dj799FMcHR1xcXHh008/5f/+7/+eeA4lUVy2a922on4ujoIOj4JiHBwcjIODA/b29vz4o+aLhnZRGxwcjLGxMSqVis/+/ZlcztbKyQqPnh607dOWxr6NAVAYK6jRswYpu1MgDZJ2JXEo/pDGiBZgNc4KqZ4ECpDyJE2yVAALcK/njrGxMZaWlgDcvHmTGTNmkJKSQmZmJkqlstR3eZ9W5nDB00Ho8yMeR59Bc8fO29u70uuzdixXV1fMzc357/v/xeqiFT7ZmvDelStXMmrUKHx8fHBwcGDq1KmsWLHiiedgCN0FrKHjIR1DDLY/KZvObuLHpT/SvH5zNkRuoHd6b0wWmNAsvhkBvgFk5WdhY2FDQ5eGbBiwgYh+EfRv2J+X33yZLUe28MfiP2g9uDUWjhZ49vLUlFF8iDHG+hfbz6MkqNqfCwfkydR3qs/gxoPZPng7Ef0iWLZgmYi2ExjUaCcnJ70+L4JGP+4aGp4fjX7w4EGpx8rIyGDTpk0MHz78iefwtAkLC+P06dOFnLulXV/qbnds26ctry96nYNOB/UiB7XRhHUd63L247P8PPhnfFx96NOgDyc+PCHn//Bx9eFLny9pca0FLzu9zMiRI6tUVM3TRjhAdGjTpg1paWmcPXsWlUrF+vXrGTp0qF6fSZMmceHCBaKjo7l06ZL8BQ1ArVYzcuRIrl27xvXr17G0tOTjjz/WO3/NmjUsX76cu3fvkpuby9y5c0ttn1qtplevXjRt2pSbN2+yd+9eFixYwO7duwEYN24c48aNIy0tjcuXLzNw4EAADhzQrGRSUlJIT0/n1VdfLfY6ubm5RERE4OTkhIODg17bsGHD5A+43bt34+PjI4tjYGAgR48e5d69e4wfP559+/aV2dves2dPZs+ebbAtNjaWJk2aoFA8ui3VpEkTYmOfPPT4wIEDuLu7yx/WkiTp3bmQJIn4+HhSU0sIGS4jTzNPRUGHR8GytjNmzOCtt96SQ0S1DpPRo0ezPmY93ft2BwWkV0+nxdIWbDLfROYnmRx2P0yTwCac/ue05kJKuF3vtiahni3QEfLUD7cPHYeUb1MeVX5Raw471XXCf40/Urok5xaxt7cnIyMDPz8/7O3tcXBwENEcLzBCnzVIkvRY+gzg6+vLuXPnKr0+a8fS4uPqQ8fsjqRfTZfbmzZtKrc3bdqUhIQEvTur5UVR+Tq0x/s06FNkPo+ykJiYSM2aNRn3n3GsOLmCYzePkbcvj7yUPObNnseyBcvITMok/pd4Lt2/xLqV60ifkc6FiRcY959xxN6N5bsl3/FP0D/09e1Lx086cuzWMYwURthZ2CHx6PNMVTDkriMo7BQoOirgNWTdLoiThRO17WpzJ+MOTpZO8pxFtJ0ADGt0wdD8F0GjH3cNDc+PRmdnZ5d6rE2bNuHi4iI7kp4nJk2aRG5uLpMmTXrisXS3sOhGDmqjCUMPhcrbKodtHkbvtb3lLY+g2U62fv16XF1dGTFiBNbW1k9s04uMcIAUQOvB3rNnDw0aNKBGjRpymyRJhIWFMX/+fBwdHbGxseHf//637M11cnKif//+WFlZYWNjwxdffMH+/fv1xh85ciT16tXD0tKSgQMHFrv/riDHjh3j3r17BAcHY2ZmRu3atQkMDJSvb2pqyqVLl0hMTESpVNKmTZsyzf2nn34iOjoaS0tLwsLC2LhxIyYm+oWC2rZty/379zl//jwRERGF7sDWrl2bqKgobt68ycCBA3F2dmbEiBF6Iv7TTz9hb28vP15//XW57ZdffjEYbgeQnp6OnZ2d3jE7OzsePHhQpnkWJD4+no8++ohvv/1WPta9e3cWLlzIvXv3uHPnjnyXIDMz84muVZCnmadCm+djxowZhRwuWjvWrl1LfHw8n/37M3oF9kJhpECtVrNv5T6ioqI0Tos4yFHlcCbpjHy+Okot7xeXXtOEWKt8VY8S62nDqveAlCZptrwokB0gKVdTiOgbwcxpM/Hw8GDOnDnyFhttfpH79++LaI4XnBddn+3t7Tlx4kSV1+eSxirYrv35ST8LKoLSOrlv375NfHw8MRtj8Lb35k7GHVq+2xKls5IJkyfoORmm75/OnV/vkJ+ZT0ZaBvtW7iP8RDjR66MhG6QsiaRdSaiiVGQmZXJ642k9B0ghWoA0XkLylQxuezE3MsfW3Bb/pv78/N7PjGg6Qi/iRUTbCbQU1GgzMzO5raprdHJyMvb29i/EGlqlUpV6LG20pq6z5HlBeyO0rFu5DOm+bkEB3chBbTQhEqw4uYLA7YGsP7OeHRd2sOb0GgZsGMCWQ1vYuHEjkovE6RqnufLgSrnO80VEOEAK4O/vz5o1a1ixYkUhYbp37x6ZmZm0aNFCFp5u3brJIW6ZmZl8+OGHeHp6Ymtry2uvvUZKSgoq1aO7Le7u7vLPVlZWZfLuXrt2jVu3bukJ36xZs0hISABg2bJlXLhwgQYNGtCqVSt++eWXMs194MCBNGvWjISEBBo1asTx48cN9vP392fx4sXs27ePvn37Fmpv06YNP/30E/fu3ePgwYMcOHCA//znP3rXSUlJkR/79u0rlX1KpZK0tDS9Y2lpaU8UAnbv3j3eeOMNxo4dy+DBg+XjX3zxBc2bN6dZs2a0bduWt99+G1NTU1xdXR/7Woao6Dtn2tC62wm3qVmzJp//63Pi4+OZOm2qXvuooFF4eHgwePBglM5KpI4Sh90OY9XXCjNrM0zyTfDz88PJ3Qn7N+yxNbPl/ebvY2tmq7mQF6AAu5Z2GCmMHu0n5+H/vwBpmr3jCjsFGAMSWFpaYmxsLL/2ugtp3co0Yn+5AIQ+p6Sk0LRp0yqvzyWNVbBd+3N5hwOXR4ReaZ3c1apVw8PDg5nTZhLZN5IRTUcQPj2cw7GHueJ9hYPOB9l1fBeBgYGMbDYSZVcllkpLbO1seX346wT4BtBsUDOwAIWFAotOFji+7KhxaHjqX8vO7OEXl4K5P+YDe9BsgfkFjE8Y4+fpx/oB63m/2fty6cXyiHgRVE1eZI12cHAgJSXluVxDl1XrjI2NS6X3N27cYP/+/YXeC88S3cTWhpJZ67bPmTMHMzMz5syZU6ZrGNJ9H1cfal+pzZYPt/DPjn/kcrThJ8IJ8A1gUvtJeNt7k52fTQ1lDSxNLFFJKi4mXWTOb3OoVasWibUSWXdmXbnlnHqRMSm5y4uFp6cn3t7e/PrrryxbtkyvzdnZGUtLS2JjY/XuPGqZN28e58+f58iRI7i7uxMdHU3z5s3LLQlUzZo18fb25uLFiwbbX3rpJdauXYtarWbz5s288847JCUlldnr6uzszI8//kirVq147733CpVQ9ff3p27dugwbNgwrK6tix2rVqhX9+vUjJiamTDYYwsfHh3nz5iFJkjynU6dO8dFHHz3WeCqVijfeeIPevXvzxRdf6LVZWlqyePFiFi9eDMDSpUtp0aIFxsbGhoZ6bCq67rg2tK6WWS3i4+MxszbDyM5Irsgy7j/j2LdyH68Pf50bN24Amg+H8BPh+Hn5EeUZxdq/1pJwK4How9H4L/dn1elVDG08lPnd5pOjymF97Hpy43JBgtTTqXDs4cX3oIn+2M+j6JCuEmatzHA840jGzgwUKGg6vCmKNxTE3o3VW1RrX5uaNWvKCVi1xwUvJkKfNfvLq7o+a8fSRXcsHx8fTp48KYeonzx5Ejc3t0L5Bp4U3UXs4+qONuquJCe3s7OzrMGAvDgev2s8G85sAAU4WToxv9t8riRfweIVC4YGDJX7AXwy5hPuN7xPSnYKV1Oukv17Nkggxem/x9PzHn5p1M39wcOfLdA4TSSw/NOSfdv2EXs3lqi4KDZEbqDzt51p9E4j2tdqz7IFy+Ry6QIBFNboY8eOyW0vikY/j2vowMDAMmmdhYUFp06dKlHvIyIiaNu2LbVr135i+8sL7bpYy6rTq9h/bT830m7oHQOYHzifqKgog1Vw4GFp2j9CSclJwd7Cnv4v9+e7Jd+R9iANS0tLMjIy5MIBAF+HfE1uRi5fh3xNDdsaBE0JIrddLknDk2hVoxXndp8jb18et/1uk90sG9Bs3TqmOEZD+4bcS7mHi5WLKD1eDogIEAMsW7aM33//vdD+KiMjIwIDAxk/fjx3794FNEkatfsHHzx4gKWlJfb29ty/f5/p06eXq12tW7fG1taW0NBQsrKyUKlUxMTEyB8wq1at4t69exgZGclZgY2NjXFxccHIyIgrV0ofMtWgQQPefPNNg15Pb29v9u/fr+eR1vLHH38QFhYmvz7nzp3j559/LnMooSH8/PwwNjZm0aJF5OTkyM6JTp06GeyvUqnIzs4mPz8ftVpNdna2XMo2LS2Nixcv0q5dO4P7JW/evMmtW7eQJInDhw8zc+bMcv99Pg20oXXu7u54eHgwZfoUPl33KQu/WAjA0ZVHUaeqObryKKC549mtRTca3mgo723XbksJDg4mwDeA7nW6czn5MsM2D6P/y/2pbV9bzveht1DIQXN30UvTRk+gBagkFV0HdMXOxo601DQOrzksVzYwdBdCNwGrSLInEPpc9fVZO9bdu3cNjjVs2DCWLVvGmTNnSE5O5quvvmLEiBFPPIeClEeEnqHtIWW526obNq2tDOPn5UfT+KasHbVWb4xNZzZx9t5ZPG09sTC2kHXZopOF3phq6eHeQ508TfLPXYGeYOFowYTJEwAI/SOUH/7+gdBZoSTcSuD3lb8zb/Y8kfhUYBCh0c+fRhvSuuI02sbGplR6HxERUSHa/CToJrY2lMzaUOLrojRbm9fj14u/suHMBj756hP2freXtNQ0cnNzSU5O5qNJH7Ht3DZi78aSp9K8frmqXKZMnUJmUib5UfkcunGIf+35F1l7s8hPySd7V7ZcjUutUKNCxcqTK9lzZQ9XU66K0uPlgHCAGKBOnTq0bNnSYFtoaCh169alTZs22Nra0qVLF86fPw9AUFAQWVlZODs706ZNG7p161audhkbG7N9+3Y5S7SzszMBAQFyYs5du3bh4+ODUqlk3LhxrFu3DgsLC6ysrPjiiy9o164d9vb2HD58uFTX+9e//sXSpUtlIdalffv2BjND29vb8/PPP9O4cWOUSiXdunWjb9++TJw4Ue6zfv16vfrgSqVSvkb37t2ZNWuWQXvMzMzYunUrERERKJVKPvvsM/z9/eU9pqtXr8bH51EEQWRkJJaWlowZM4aDBw9iaWkpL0K3bNlCRkYGy5cv17Pj+vXrAFy+fJm2bdtibW3N8OHDmT17Nm+88YZBu55mItPSoGuPNly5mls1bty4wbQJ0+Tw5di7seTk5wCQnpYuJ0LV5gDRhgS6dnDlnbB3aNunLQB/3/qbXy/+ypqVaxjaYSj2sfYoWijgM5C66NypkdDcUTwN7Ye2x3+kP71e6sV7jd9jUvtJBAcH41bdjTbvtZFLKY4ZM0av3C5ovkAsWbJEJNkTAEKftVRmfba3t+f//u//2Lp162Pps3aspKQkg2N169aNiRMn8vrrr+Pp6Ymnp2eFOKi1zou2fdrKIdFlRTecWsuUqVOIj49n/Ofj9T47DPX1cfUhol8Ek9pNYvJvk/nvsf/y2e7P+Gv1XyTcStDT6pScFNSoSchMYE7XOdi2sUXxmQJFS4Xedhc5H0gLNHmaQBMF0vHhsRbgNd2LE9VPUGdBHfbF7UONGvs37VHYKfDo6cGEyROEJgsMIjRaw/Ok0YGBgcyePZsFCxbI/YvTaIVCUazeA/z111/Ex8c/9fK3Ja3Jdbfx6SazDvANIPSPUEIPhRLgG8Cf2/6kZs2aJCYm6hUF6DKui7xVJikria61u9LjpR50rd2Vezvvada+RtCuRzsUtgry2ufx2a7PCD0UitRVkh3NVl2scKvuRs1eNbmRdkPjHNE6okGvGpcRRoxsPlJ2hpd31bEXEqkK0KJFizL1DwkJKVP/ffv2lan/88zzMlcPDw8JkDw8PB57jPKaqyFbyvoeK08K2rN06VJp0aJF0tKlS/X6Be0Mkkz6mEgokABJ6ayUQuaGSEpnpWTd31pq/kNzyWaWjWTS20TCFsm4t7HkMddDUvRSSNgiYaE5z8bZRlJ+pZSYhkSvh8eNNG3ah6mDaant1j7s7e312pcuXSp5eHgUmkdBnpf3cGkoq7ZVVsoyj8f526lKv/OSEHN9csr6HgvaGSQ5z3GWgnYGlaq/rlYZOrfmkJoaTbUylrV6w4YNktJZKZk1N5OUzkqDem0200xSTFNIRtOMJJPeJpKJvYlk0ttEevm7lyXnOc6S/2Z/6eXFL0umM0wlj3kekvVX1pJimkKynWUr6zUWaLRa92H7sM320bFea3pJxtONJaYhKaYppJcXvyxtPbtVCtoZJMUkxJTp9StIVXoPVwWNrsg1dFX6XZeEmGv58SRr6Mf9fhC0M0iymGkhWXxlIQXtDJLH0a6fjY01em1kZyQF7QwqpO1BO4Mk6/7WkqmDqbyGpgkSCiTLFpaSW6ibxDQksxlmkvd8bylgW4Dk8LWD1OrHVlLt+bUls5lmUqsfWmnW2BZIRpZG0isfviL5b/Z/Ys0tLVXtPVyctokIEMFzSWUqu1eZbIm9G4tHTw+snKwYFTQK0Oxjz83NZcaMGXp3GP28/Hip60s0eb8JVk5WSB0lrnhfoe/SvnTq34mMPzN4MPsB+f/LhzRQRamI3x+P9Iuk8UwD2EJW2yyyjmbBbGA7kI1c3QUFmDqYEhBUsrda+zpq98Tm5ubqefGfZsUcgUAg0BLgG2Bwy0lRdxp1tcrQud99+R3N5zTng0kfyJ8dt2/fJj0xnbyTeaQnphfSuQDfAHxv+mK+yJxGNxoxePhg2s9vj9RCwk3pxtDGQ+n/cn+S/0gmb24e8VHxZORnICGRmVdC9TLdrTBo7jZ+3flrOtTqgKmRKUYKI1pWa1luJX8FAoGgonjcNbnudsMA3wB5nGrVqsmRyG7V3eSk0wG+AXSv253D2w7jXsMdu1g7+r3Xj27/7Ua/If2oH1cfTgESZJ3IIiFLk2w3V51LI7dGrDq1iuScZI7dPka7Wu0Y23Isd5LvIO2XNNW8TCVcO7oyqd0kLt2/hO+Pvmw7t62Q3ZUtCv15QThABM8llansXkXbUhZxCz8RTrRHNOogNak+mrBObcm34OBgOflT+IlwouKiuJd5j07vdOLomaMEBgSCArZf2M6+Tfu4EHnhkaPDAshFk9j0YUJTugIdIT8qH/VOtcbxgabNpJYJKKCBXwNy7+fyffD3so2GMnDH3o0lMDCQXcd38WrAq7hVd8PMzEzP4VGZHE0CgeDFwcfVh/Obz5NwK0HPMVGUU1ZXqwyd26dBH058eILvg7+XPzu0VWCGvDdET+fCwsJwr+HOuP+M4+KWi2Tfz+bS1kvUvlqbo58fRXFcQU3bmgT4BjB9/3QSfk3QT2zKw5wfXXmU46Mg2q0wLTSVuia1m4SPqw/N3JthbWZNE7cmTGo/qdjXSCzCBQJBZUC7JgfKpEna7YYRfSPwcfUpNA7AnZt3+G3hb/L2GSdLJ46uPUrCrQSWLVgGCthzZQ8AP4f9LI+taKzAzdINE4UJChT8c/sfclW5ALhbubP57GYij0eSdygPHvqrJS+JXy78QuihUKbvn070nWim7y+81VPcHHw8hANEIKjklEXcCnqwQfNh0LhxYwIDAw0mf9LuddwYuJHaV2rjbe9Nxp4M2dHh3NsZEyMTjYNDhcYZYg5mRmZyFQEpTyf3R0/Ifz8fRYgCr6ZehT6AdJ0wuj9r2056nGTwssHMmTNH74tAZXJ6CQSCFwtDDtiinLIFtao4563WCWyqNOWdsHeYPG+yfG5YWBhjxowh4VYC+1buw6qLlRx1FzorlMykTIwPGvP37b+ZsncKV1OuYt3V+pHD+mEVTjVq/XwfumXKAQdzB2ooa2BmbMbQxkOZ1UWTQyDAN4ARTUcQ2TdSjvooTdSLQCAQPGt0Nako3SqN4/b27dt62qZ74y7AN4DXh7+OW3U3RgWN4u+bf6NWq0nJTsG2q60mCfXbFpi8Y0KuOhcvOy8kJLLysuRcTHcO3iFjTgZJh5O4c+AO5D+8cJzGIY0EbRLaYLLAhDYJhZPhaj9fRgWNeuxcVS8iwgEiEJQT5X0HTDuen59fqSMffFx9iOgbQUS/CDnR6fhd48nOz5bbA3wDZIeDtoTiZ//+jPj4eJYtWEabhDYo8hRYKC1wH+ROo+6NHpWhk9BUd8mGvH0PEzZpC78ogF5oFtpoEu3t+r9dxMfHMyV4imxjUU6Ygm3C4SEQCCoLhvSotBplqJ9WmwNCAljYbyGXzl5i8Q+LmbJ3iryInTFjBiqVCiMjI14f/jr/+uRfOExxwLSVKSavm4AF5GTlcHb3WWLuxuBo7ohdWzvMLc01Duv9PEp+uvnh/3uQI0RMjUxpcqMJZt+ZEUAAY1uO1Yv00E0WqKU0US8CgUDwtChq7a2rSQV1S3vOxIkTCyXeL0i1atVwq+5G/X71ib0bW+jGXePujflx74/8ZP4Tl5MvU8exDg9yHnC3wV1sJ9lCC8hT55Gck8yllEuA5nlHz444mDs8Kkd+AEz8NLqOBdi/Yc9rnq9x7NYxIuZFkJeSx7oF6wrZp/18SfVJ1bNLUDzCASIQlBPlfQdMO15UVFSpFtlaQZ82b5q8gNYKdWJmotzPUNRFXvs8jOyM6BXYi8jvIlFnqck1ziWpYRL7r+1H1UWlCZ82QY4MkTpKGmfHW2jaGiMvuM2NzDUXe7i33Krzo1r3hjJw+7j66JXgFXvMBQLB80ZZtyuuOr2KUxtOIWVJoIb8fflcTLoo63NwcDAODg7Y2tpS7X41/t3z32QfzcbH1QejFkZgBmSD4oCCRq6NuJZ2jZsPbmLayRQjS6NH2xbT0OxFf7il0cTeBKduTnSr042EnQly+HZp8nuUNupFIBAIngZFrb11NamgbmnPUSgUGBsbo1Kpily7Ozs7M3jZYE56nCT8RLjejTqtjk/fP53LyZfJV+eTnZfNidsnUKPGydoJG3MbeSwFCsyNzPGt5ksN2xrk5+Xj+porxnbG1Otbj/Z922P6b1OYDNU6ViM1J5XzSefJyssCeHQz0gBau+xi7cR2xFIgHCACQTlR3nfADI1X3AJbK+jzZs+TF9BaQXS2cpb7GYq6sG1ri/o1NeHzw/F91RcsQJ2jRvpb0oTgacOntfvIez4c7Fs0kR+fAXHIdxadrZ3xtvem0RuNcPi3Az4uPti42DBt3rQi51vUh5jYWy4QCJ4HyrpdcWjjoXR6vRMKhQKFQkHnEZ2Z3WU23et2JykzibZ9NKXYU1JSWLNmDemJ6eTvzuf8F+dpe7ctRn5GYAFSjkTUpijsze1RHFeQviddk/dDJy+T/H89yFfnk5mXycEbB2n0TqNCOUeK01vh6BAIBJWJ0qy9i9qWGBoaypIlS0pcaxfcMr521FrG/Wccfl5+DG08lJCOIQzyGUR9p/rcSLtBWm4apkamdPbujLmxOeZG5lgaW9LRsyMDfQYSczeGuKtxNFQ1ZGHQQvJT8unxbg9i7sXQrW43mrs35+suXxPSMYRm7s34cPKHeHh4MGfOnCLnqL2huGzBMrEdsRQopOLcSc8JLVu25O+//y51/wULFpCSklLq/vXr15frlFd1xFzLB3t7e4KCgsp93Jo1axIfH4+Hh4ecnElLWFgYM2bMYFTQKFJ9UgnwDZDv5kVFRXHx4kVmzJhBcHBwocXrtnPbGPDqAPJS8jCxNyFfran8orBVYPovU/KO5mkyUz+sEsB+NIma8tGE63VFc6eRhz9fA06DQysHjN8xJnVWKnkpeSidlXw761s9O7R2+/n5ERUVVci+4uZsiKioKPz8/Mr82lZGyqptlZWyzKOs+gxCt6oqFTXXitJnrZYZ0lhDxN6NpXXD1mQmZbJw4UI+/fRTALqM68K+lfuo51uPO//cQS2pMa5vTMr5FBS5CtRZaqycrLD43IL7X9/XOJ5t0Tiiv0Xz3BSNPjcGPNFodkcehVvbQvM5zeX8HrF3Ywn9I5RNH2wiMymz1Hr7OAiNrlxU5Bpa6FXVpKLnWlqNLqvmlhbddWdkZKSeXrnXcCfhVgIKOwXj1o2Tt5KDRtM7/F8HknOSAbA1syUtNw0FCkyNTfFx8eGLdl+w9H9LqfugLqN6jcLX11c+V3vj8kmioJ/kNalK2gwlaNtTKsVboVR0HfaqVhe5OMRcKzdLly6VPDw8pKVLl+odj0mIkYJ2BhVZK3zfvn2SW3U3CZDcqrsZHLPnOz0lE3sTiV5oHrZIJn1MJO/53hK2SKCpZ47Fw5+1DxPkdmM7Y8l8prmmH0hGRpp66WOmj5GUzkqp5zs95VrqDg4OkoeHh+Tg4FBszfai5lwUz+PvtSgqWtueFkKjyw8x18pPUZpV8HjQziDJpLeJhC3SjxE/yv20Wm1kZCQBktJZKdnPtpea/9Bc1tImo5pITHuk1fRC/7lWp20fHp+m327Z11Lv8yJoZ5BkMdNCMuljIimdlaXW28fhef29GqIqaHRFzqEq/a5LQsz16ePh4VHs+vFx0dVq7VxD5oZISmel5P2at6SwU0g1h9TU01Bte8tuLSWFnUJS9FZIrqGuEtOQjHsba7S4iUZ/zXqZSdYzrSWT6SZSo+8aSX7L/aSXF78sbT27VR4vJiFG6vxpZ8mtuluF6rEuleX3Wl4Up21iC4xA8BxRVPhxwbwehmj0TiOM7IxwqO+gF9qnm2vEeIKxZrvLwy0v+c3zuZF2A4c3HDTh0xKYYoqpnakmHwhgbWVNm/faYOVkxQeffUD7e+2xsLBAoVDQqlUrNgZupHm15jy494Dow9GoVCoUCgUpKSnEx8cjSRIeHh74+fmJrS4CgeC5p6itMAWPB/gGMHj4YNq814bs5GxZ+2ZOm4mHhwfvvfceHh4eTJg8gRFNRxDSMQTz1uYcjj2MY3tHTWLTXwAv5OTThbYrdkSfh+09BvXg0v1L+P7oy7Zz2+QKYoOHD+Zw7OFCnzG6lQ8EAoHgcSjvLc1PuvW8oK5pn7t2cOX1Ra/zfyf+j3+i/8HE3ITpn08nPTGdW6dv0Sy0Gd99+R0AwzYPw8vfi+n/0rT//b+/kVIljA8YY2ZihqmRKaoolSby7jSQBrn7c8lQZZAv5ROTFEPUtSjOJp7VK3MbfiKcfSv3FSq/LigfhANEUOWo6jkjDC1EtfsT/bz86L2mN7UX1qb3mt56fT4Z8wn1vqpH3Mk44uPjGTtxLK+veB2Pnh64VXdjwuQJvFHnDUy0no2H5Ev5pDdO1+T9sIWab9dE/ZkaugMmkPEgg7rpdclIzOD74O85v/k82VnZ1KhRg5s3b+qVIUtPT8fBwQE7OzskScLY2Jg5c+YQHBzM2rVrS/WlQSAQCMpCWb68F/f5oW3z9/cv1Ef3GkUtykcFjULprGRU0CjgYdWufhHE/xJPXl4eY8aMYdq8aZypeYZdx3cRGRnJjRs3mDZhGvO7zScqLooVJ1fgv8WfmJ0xmsSmEppFdUG0jpAWBtqAHRd3MGXvFP658w8jt43k0v1LOFk5MandJIPh16VxsgsEAkFxlPd67klzEo37zzgWvbuIcf8ZB6CX1HTDmQ0cXnMYtUqNKlelOUEBDm86cCPtBlFxUYSfCGfDmQ1c+/maXCCgxqs1UDoroSPcybiDkcJILghAY83/Nl1tsDO1w0RhQiOnRvh5+vGy88uEdAyRbdMtsfv/7J15XFT1/v+fZ2aAAUYYBAVtKEHNBVPB7NrqmFpaodfUW5ZY3eDevLcS63cjW9i0XHJL/aZXMC1NK3OhTcu8YqtlmiW4i5aYoiCIiGwzn98f4znOwAyboIDn6YMHzpnPOefzmXN4nfe8P+/P+61W12p4VAeISoujoQW2qTlUnBmicvKj9KPpfH7wc44UHOHzg58z/dvpTNw4kRM5Jxh751gObTpExZ0VSL4SFXdWkP57OrtMuxizZAyjo0ZzMO8gFVRcKp24w/ZTMasCSUjwHLjf7s6d19+Jz60+YAEErFq1CsDByREfH69UMSgqKuKFF16goKCAwsJCHnjgAUwmEwsXLiQmJoYXXnhBiQypLPRqeUUVFZXLoS5f3is/P+wdG/J7K1euJDs7m1cTX3V6jpiYGDbu2Mie4D1knspUniFfHPqCitgKskKzHM4pa5vFYmHWtFks+3UZo1ePZtiqYbaflcPoFd2LRY8uQrdTR8apDM58cUbZX7pJoibkJHwGNwM6ScdNbW8i0DsQgIKSAiZtnlTtZ1Q5ebaKiopKXWlq9lzGRxlYz1rJ+CgDuKRzCf0TGN19NP0e6YdGq0HrrsXbx5uBzwxkUcIiZcJxW9o2xGyB1EFSCgT4jPHh4SUPI0VIaNFSaimFPiD1l5B+lzDea6TPA30Y1nUYu57axe6nd7Pl8S3s+fcehncdrvQtrG0YX735FZMTJ/Nq4qvcOv5Wxq0b59KR39S+qzR1dDU3UVFpXsg1vxtCYFNSUhg/frxSIqspZL6XDVBzBzMTN050SJgUHRHN4TOHyTidQY82PUCCZb8uI9gjmKLcIrRbtbSLb0f57eWcKj5FW8+2tPZujbmDmenfTmd/3sWkVnKivK22l6JQ4PatG4Y7DPyr77/Iys8i43QGXc1dOfj1QYKDg9HpdHh4eFBcbEugJ39W8pcGPz8/JEnCYrHw6aefkp+fr4xJkmwGvNFobBKfsYqKSsvBvuJVdaSkpHD+/HmMRqPy/JAdG3Dp2WLobODAzgP0GNXD5Tns9/so+SOys7PJ/TCX8mfLKbhQ4HDemJgYPvroI4xGI8VFxZxPOk9B5wL2Ht1rmznsA6wGCqH0i1I0YbYZRe3XWrR3aSmLKKvxMyi1ltr+Y4Gu/l05cvYIQzsOpat/V7IKsujk14l7O97r8jOSnewqKioq9SUmJqZJ2XiTEyc7fF+QdS7zVCb+nv7EJcWx9+u9tGvbjidjnyQrNIs1e9YQeiSUsVFjOX/+vK2MuQU8/+OJRtJw6MwhDnx5wLbsRdZvgK02W7rgiwLSw9Jx17iTcTpDSUTtiuTkZHL+zCFnZQ67TLvw9/R3qsX2zvum9Bk3VdQIEJUWR0OW6UtOTsZisaDVapuMx9o+2sNZJMjHj3xM1oQsPn7kY+JujyPEGIJoJcAHrHdZOXHuBPkl+QR4BXBL8C1kfZXFiIgRvDf2PcSOi0WhLobrSf0l2//1UF5STv53+cz4dgahfqEE+wQz7a1pVFRU8Pvvv2OxWCguLlbC9WRvtNlsVsqNGY1G4JLDA2xfOoQQ+Pn5MX36dIftwcHBvPDCC+oSGBUVlXoja2ZNmfWTk5PJz8/HYDAozw/7yAf52fLR+x/x7PvP8ubLb7o8h/1+8qxnux7tsMy2cPqb04DjjF1AQAAGg4GykjKbQX1xrbi0VbKVIpdDqO+CG1vfSLv+7bDEWiizll2K1gPH6L2L+Ov98dZ5K69bubci2CeYkd1H8tHfPuJfN/+LqYOm1uozUlFRUbna1CbawT56z7595qlMxq0dx7h147ht+G015tU7ceIE2dnZvP7a63yY+SEf7PmA16e8TlFuEW4aN7S+WugPFywXOP/jecpnlsNXOEwiAmgHaNEatfgP8cd8g5mOfh05UnCkxsjE+Ph4AtsH0u+RfozuPtqlk7qpRdc0dVQHiIpKNciCIi/VuBwuNzxNFvO0fWlM3DhRqT9uL4aZpzIZtmoY3RZ0Y/6P80ndmUpC/wR8/XwJfCUQ91vccde646f3I/9CPl8d/orS/5UiLgisF6yXxPri+nHRR9hC99wlKAG2QnZRNkt3LVXWQALo9Xrbfho4X3YecEyuKj9g7r//frRaLffff7/S5+TkZAoKCvD29nb4jOX9JUlSRV1FRaVRqbx8TyasbRjREdGk7kxVQo/tnR2udN2Z0+XETyegEH78749ERUUxfvx4B+dufHw8rXxb4e7tjneEN/iA6CDQzNUQ4BmA2/9zo+u9XVn9t9XoLDrbmvOvcTS0K0XvAdwUeBNuOjf8PPxw17pTVFak6HdtnUMqKioqTYXaLHW3d2LYt5fzdqzes1pxPlTW8eiIaIZ2GkpecR4BbQMwBBjQmXWE+oUSagxFdBBIGol7ht7D9UnXI/W56Ki+qL8WqwWdjw73/u5Kf1rd2gqvF7y4/u7rWXDfAlb/bTWP93q8xsjEmJgYTh4/yd8j/s6WZ7fwfdr3Lts11OTvtYDqAFG55qnOMVGdoNTVoXG5uUnskzOt2L2C+Qvns+rJVUx4bYJimE//bjqfHPiEfV/u49l7nuW/i//Lmj1rKLWUklOcQ6mllAuWC1youECFqOCC5YIS4YGeqhUDLiL6C6WigPl6M/1y+lEyowTfTF8A3nzzTQLbB6LVaSnKteX7cOaNTk9Px2KxkJ6ermxz5bWWt0+fPl0VdRUVFYXGWOvsyhELjoZ05XPHxcWRnZ1NXFycwz7OcoeUltqWoQghWLlyJRaLBY1Gw5OxT3K04Cjf+H/DDwd+4LYFt3E+8rwtielRsBRYOPPFGQK9AunUuhNp36RRer7UVplLjgy5qN1+9/oh+UgOWn6y6CSP93qcpAFJhLUJY3zf8Wo+DxUVlWZLbaIdfDN9KZlRQulPpXR5sIsSnSxXvLKPpqhsn4e1DePP9D957+/vca7sHNsyt/HPf/yT1aNX89HfPkKfrUdYBf/b8j+OFR5Dp9E5RuoNhornKtDdYss0ISFxR/AdGDON7IrbxYTXJjgsK6xNkm61IEDDojpAVK556isqdd3vcsPT7JMzjb1pLLs+2EXOnzn8753/XQqhu7iCRfZCl/6vlLT9aZRUlCjH0Uk6LFZbRmsJCbeb3fB42QOfeB+kPralKe644+9hC5vu264vPe7tgcd/PJj07CS2PLGFT1I+oSi3iCVzlwCXPNQ+3j4AFBcXM378eMxms8OXCWefgSsnU1282WryJxWVa4fGSHTtLPpDxn45S+VzC2ET3YKCAgf9sXeayLp3yy23IEkSnp6e3Nj/RjS+GgY8PYCzYWc5c+EMK3evZPq30/nx4x9hGrafVoAE1husZBdls/HgRmbvns1J6SQAbn3d8H/JX1lnfjbsLJ4veBJ+fziTbp9EG682/Kvvv5gzZA5Z+VkcKzxGVn6WS8M7bV+aUhpXRUVF5UpQVxuuNvbhotmLKMot4p157/Cr6VfGLBnDbcNvI3VnKnF3xPHuiHeVyDdntqmcIPX8GVtUs6yZox4exfm887jr3RF3CSxWC+XWcqxYq1TfKrYUE2IMYexNY5k6aCpnPz2LOCvYtmybor21TdKtLnFpWFQHiMo1SUpKCrt37yYlJaXeolLX/S43PE32Fg/vOpw5Q+bQ+6HeSL4SpgdMihc77o44+rbrq3ih3Qe4U1hWCECgVyB+ej8AzlfYBF0gKBfllFpKKSwrRFz0oJRRRl5pHtYdVjJfyiT321zKLGVsPLyRzFOZDt50e6ZPn47JZKKsrAyLxcJ7773XoJ+BK1TPuIrKtUNDG4LVRX/AJe39Pu37KklSZ8yYgVarRQjhUDnmcP5hAjwDMHcwK7p3/PhxhBCUlZXx0P0PKXlEfDN9IQfETsGmjzZRklZiW3JYAvyJzbF91NYXjVVDO0M7pW/l1nLKreXoJB0aNFix4u3uzfIRy/ns/c84PeU0qSk2w7pyJRdnhvczU57hlxd+4ZkpzzTIZ6uioqJSE3Wx4WrrLOkxqgcaXw09R/dUdK+y5smRenIukLZ3tlUcwJMTJ9vK2baylcsNui6Ioc8NZV/6PpuOl5RxYfMF9L/q0aJ12Y/zZef5+cTPHDpz6JKNbSlT+lFThS15vECDRaSrAKIF0KdPn0Y9/pYtWxr1+E2Ja2WsJpNJzJw5U5hMpqvdlXqTkZMhYjfEioycjCrvRa2NEvopemFeahZ+0/zEzJUzhVuym9AmaYX3a96CRC79RCLwufg7sdKPDwIQHn4eouv8rsK81Cy6zu8qvKZ4CZ/XfUTonFCxfu/6Kv3SumsFIDw9Pa/IZ7F48WJhMpnE4sWLW9Q93NjadqVQNbrhUMfa8NjrR3WYTCYBVHluVN4/dkOs0E/RC/1kvYjdEOvQDps7QwBKe/l5pPXSCqRL7yMhJD9JICG0PbVCn6AXukSd8P+bv5B8pEuaHYmQfCUxJHaICF8UrmiyzqgTgNAZdU7HIz9DEmYmKP33D/QXgPD28a7VZ1IfWtI93BI0ujHH0JKudU2oY60/tdVgIVzrcGWc2ciVt8VuiBUBMwJE7IZYsXjxYuFmdBNEIsIXhSvtl6ctF4HtAwXYdJaeOOi45CMJbaK2ql2tv/hzUadD5oaIrgO6CiSE/nq98PL3EuOTxju14+0/j9qMt7afSU20tHu4Om1TI0BUrkni4+Nxd3dv1qFkrpLXZZ7K5NfPfqX8jXJ+/uRnvNy80KCh3FqORVjQSpU81U6S5sno79aDD4Q/FM6QTkPYdnwbWQVZWIWVwrJCss5mkbQ1yWGf1J2p6CP1GAIMvPnmm1UPWgdq69VWkz+pqKjUl8r64Up3aht5Yr/G3DfTVzlWZX2yT37q5uaGRtLYTGq5SJYAztp+a49qudHnRiqoIO/jPEShgE22yD42gTgr2LR4E8tHLGd41+G2cU2Mwc3PjZiJ1eviotmLlNnXqZOnYjKZcNe6q1F1KioqV4S62HC11eHv077no5iPHJKGVletKzk5mfKCcty+dSOhf4LS3uRj4qnnnsLL34vA+wLhQWx5PgAkW448CxbHk29FieKTttoEXRIS+fvzQUDJsRKK84pZPn+50+Uv9hExtRmvujym7qgOEJUWR22+NMfExHDTTTc1my/M9gn1nGE/5tSdqfy2+jcsZy0UfVXE8XPHlbA7CYkSS4njzpWS6NkzePRgYj+I5cUJL7ItbRvMhojjEcwYPIMQYwihvqH0y+lXJXt2THQM2zK3Xfbnqy5tUVFRudK40p2YmBji4+NJTk52eL68mvgq2dnZTIqfxMSNEwF4d8S73Jl3J5NfmEx2djavJr4KgJeXl7Lfk7FPkpKSQnJyMno/Pe5D3NEZdeBxsYEGIvpF4O3jTVD3IPZP2e9Q3hbgQsUFxaltsVocQrs9bvHgl32/8Fb8W07HKYeD9xjVQzGe5S8i8nJG1aBWUVG5ktRkw982/DZGpYzituG3OX0/cVYi3gHePD3x6So6XrkcburOVMwdzKTuTMW3iy9I0OsvvRQnsszOdjspeaYEES7w1HlespsfAM8+ngRIAZca7wDKQHKT8PH14aF/PUQbrzbE3hrLg/98EJ1Rx3W3XUdg+0Cinoki2CcYcwezQ//MZnMVTa7OnlYnAeuO6gBRaXG0xC/NNSVJsh+zuYOZoPuCcPNzw3iPkRBjCIGGQNp4tcFT54nFUslTXSlpE4AGDeYbzEwdNJU5Q+aQfjSdn1b9RMmZEranbEf/m56sCVkcjj3MJymfVMme3VBlFZ15tdW1jioqKo1JdbNpzp4v8lpzr4FeDjqdnJyM1WoFILcgl8RZicydOxdDgAHvkd6cDTurHK8kv4SY6BiGvjUUabAEPqC5X0PZ4DLax7fn2N5jlOaX2mYWB6NUGigsK6TN8Db4B/kT9NcgDucfdkisN/276S6d59ER0QztOJT25vZs3LHRwXhWDWoVFZXGoCYbzl5jK0/+ZZ7KJGpdFMt+XebSHp41bRbFecWUWcowBBgcdFw+duykWIatHMZbP79FdEI0cx+aS+Z3mSDg5y9+5vqx15O2L420fWlkns7ky6wvsWIlpzjHoaiAJ54M9xluq6IoczH6Q2fQcbbgLEH9gxAIsvKz2Ba4DUushbZRbTl5/CQet3goJcnt+7dq1SrF+aHSOKgOEJUWR2OFgl3NL95ymJ65g7mKMVs5KWn60XQqIir498p/k786n6wJWfh7+tPP1I/rfa+nR9seVY4vJ3Dy1HrirnHn0ZseZcvjWxzCBAc8NgCNRoPFYnEw/u0/74b+jJwZ4S3RwaWiotJ0sNcdewPcVQLoN19+kwGPDaD4q2J6ZfdSktnFx8fbyoN7abEUW5g1bRYxMRej46Jj8M305fz587TybYW70Z284jyeDH+Svu374ia50U60o8K3gkNFh9B30qPRaLjupuvwudUH3fM6dDfbSi/2HdaXqKVRFIQVsClrk0NiPQQunedhbcPw9/Jnw6ENNVYgUFFRUWkIarLh7G3KypN/qTtTOVJwhBBjiMukoc+/+Dxe/l70e7xflUjk+Ph4DAEGym4v4+jZo5T/VE7u6lzbMnCwLT8UcGzlMZ6Z8gxJW5MoqSih1FKqHMNP76csHb+w9QJflX9FqaUUPw8/+gb1xXivEa1RS3SsrX/2y2wS+ifQO6i3ssSmcgLU+Ph4tFptFTtbpeFRHSAqLY6WWGlEjqpIP5rOsl+XEbUuirR9aQxbNYw7376TH9r+gNd/vJhTNgdPnSdtvNpw+Mxh0valMW7tOA7kHeCTA5+wP28/v536rcrx5fWLgzoO4qGwh0DCwckS1jaMr978ikWLFmEymbjuuuvQ6XRERUU5fN5X4jNS1zqqqKjINLZj2t4An/7ddL5r+x33LLjHIV/IkD5D2PHuDvJy8ti/dj+AUllgcuJkvNy88DB4YDabCQ4O5vu075kzZA5L5i4hPz+fEk0JxW7FrN6zms1Zmzn44UHKz5ZT/lM5J8tOIhBcOHgBq9XKiR9O0OFwB7zcbEtpAvcF8v3E79mWto3BoYMZ3X000RHRyjMj7o64aisM1FSBQEVFRaUhqcmGs7cpK+tTdEQ0j/d6nOUjlruMMk58PpHzuef5YeEPgE2LE2clKpVUtmVuY8xjY7j/xvvx/N6TiyvE0ev0uPd2V5wgJ9edpE+7PlWOf6bkjMPSca2kxXyDmaV/XYrPHh88vvNg4YyFvBX/FpmnMpm0eRIbD23k0JlDDO86nJ3/3KkssakcMR0TE8PChQuviI17zUdTX8FkrEIIITZs2CBuvPFG0bFjRzF16lSnbbZs2SJ69eolunfvLu66664aj6lWGGg41LG6pi5Zqi/3GOv3rhfd5ncTkSsjxfq960Xshlixfu96EflepPB53UfoknSi2/xuQoqUnFZw0SZphZQoCV2STpCImLlypvKeIdFQtdrLxR/TLJMInRMqjNOMSmZqZxmqtVpblRetVuswpsv9jBriM25J9/CVrjDQGPoshKrRDYk6VkcuJ/u9vd7Ya13l7QOfHSgC2weKkLtCBD6IkLtChMlkEgkzE4QhwCAA4efnJ0wmkxg7dqwwBBiE90hvEbshVqke4B/kr7T1D/IXwY8GC+9W3sLD4CE8R3iK+R/MF13mdRH3zbxP+ET6CGOAUSTMTBDmpWbhkewh3P/qLiSNZKs84GPTa02iRrh7u9u26RGeIzyFf6C/U/2srmrYlaYl3cNqFZjqaUnXuibUsTY95CovsvbKzwlZ48eOHSsC2wcKD4OHUnlLrmKIHuGR7OFgP5OIIKFqBcXwReEicmWkUjkxsH2giFoTJULmhggpURIkOlaWcfasuRzqcxxnz87mcl1rS3XadkUdIBUVFSI0NFQcPnxYlJaWip49e4rMzEyHNvn5+aJbt27i999/F0IIkZOTU+NxVeO64VDH2rjU1lgPXxSuODLCF4UL4zSjaDOjjXCf7C40SRpBIqLvf/sKra/WwSCWfzSJGofXM1fOdBDrrvO7Vlv+NnROqCLScokwe8aOHSu0Wq0YO3as0zHVV9QbopRXS7qHr6Rx3Vj6LISq0Q2JOlZHLseAtNcbe62z37548WLF4avRaBx+u3u7C8lTEu7e7g4lbQFhCDAozhONr0YEPxosdL11Agnhc7OPYijLzpCZs2cK9AhtpK2cos9rPiJ0bqhwS3YT+r/qhb61Xjww6gHh5e9lK3ubKInQOaFC63nJWJeP6Uw/XWn51aAl3cMtwUmtOkAaBnWsTQ9npb6FuKTTRqNRGAIMijZzsQS4l7+XCHo4SOiSdI728wN2tvJFvdUZdWL93vWKXa0z6sTAZwcK/WS90CRqhCZRI/ym+SnlyV09ay6H+hzH2bOzuVzX2tJkyuD+9NNPdOrUidDQUNzd3Xn44YdJS0tzaLNy5UoefPBBrr/+egDatm17Jbuocg0gh33l5uZe8XPXdvlGQv8Euvl3477O95HQP4EQYwilllJCjaH4ZfrBbMjanIXUX0LfWk/P0T1p69UWnaQDQOwQMBvHigF25W5zinIcXhvcDJivN+OmcQPAw82DsLZhLsOjly9fTkVFBXfddRfnz5/HaDQ6TTRV16Uw6vKWq4eqzyrNkZiYGDbu2Mie4D0uq2S5wl5v7LXOfntcXBwWiwVJknjkkUcwmUzKbzetG+KCwMvbSzm/2WxGq9Xy1yF/JaxtGG++/CaPvv0onjpPKn6tAAFFB4vwHOiJt783pbeXkv1JNliBEhBf2+KxC8sLySrIotxaTsn/Sig5U8Kubbv4ac9PRI6JpKt/V643Xo9lkMWWMPUeDf5D/PEP8neqn+pSl+aPxWLh3//+Nxs2bGDPnj2sWrWKPXv2OLQpKCjgX//6Fx9//DGZmZmsXr36KvVWReXKUrnCi1yRKzoimrNhZx0SPUfGROLm50a5tZyi3CKk3yU0PTUggUc3D0qfLUVECGWpoWIvf217KSGhv1uPm58b3Ud2Z83eNfyr778I9gnGw+LB9ne20/t4b+6/8X4e7fko3zzxjbLsxdWz5nKoz3Gu9UTXtXaA/PLLLzz44IMEBASg0+nYuXMnAC+99BIbN26s1TGOHz+urMECMJlMHD9+3KHNgQMHyM/Px2w206dPH959993adlFFpVbIX85PnDhxxc9dW8EZ3nU4e57ew8djPmZ41+EsH7Gcv/f+O9MGTePC5gtQCIWbCrFEWOg3qx93j7qbMmsZQtiMZ7FJ2MR6AzZHSD5Qhi1TdX9o5dEK9wHuSL4SrQa3oo1XGzr5d0Kn0aGTdNxw6AaCg4NZvbx64yk5OZn8/HwMBkOVRFP1EfVrXZAvh8vVaFWfVZorNVXJcoW93shrsb9P+57k5GQlA7+sqb6+vixfvpxjx46xfPly4uPjKbeUI3naNFTOzbTpf5uwWCx8+umnSr4Pfy9/Dqw7YFtrLoH1LiulvUqx3mWl6LMixAUBEui8dLS+tzUSkq2DO7Dpdyvby9zcXFYvX82hvENkFWTx+9nf0d2sw+P/eeDdz5veQb2xllp54YUXqqzrbsjqXCp1pyFsaNVJraJyico5LOwn3uyfCfL/J301iYj/RpC2L41tgduomFBBq/tbYQgwEB0bjTgqQMCZvWewCAs5xTkUVxSjRUvwXcG2nB+tgCTQf6LH2seKdYKVzOszWb1nNVn5WYitgvPnzlN4tpDsT7P5eMzHvDviXQfdtdfihrJ5Vdu57uhq0+jbb79l0KBBhIaG8sgjj7BgwQLlPY1Gw6JFixgyZEiNx5ENCXskSXJ4XVFRwY4dO9i8eTMXLlzg1ltvpV+/ftx4440O7RYvXszixYsByM7OJj09vTZDqRdFRUWNevymxLUw1jlz5nDixAlMJtNVH+uRI0c4c+YMrVu3JiQkpNq2w/XDyT6QzZQZU+AcaH20lOvL0ev0dNR15Po212MttIIBeA0lsROAyd/EzCkzQQsEXtz4kO1HQkIg4DxMDp0MgJQjMSh2EBqdBr+zfvyy7RdO+5x26E9ubi7PP/+87fiVPsvOnTuzfPlyAIftubm5nDhxgnbt2hEQEEBjcC3cw5VpCI1uSH0GVaMbC3WsVblXdy99uvQhQBdw2Z9NSUkJsbGxlJSUkJ6eTkpKiqJZ8rFLKkooKi7i9aTX0eg0tApuRamllHJLOV7TvCgtKMVisWCxWCgsLCQ8P5y5c+ZSUVgB7tic0e7A7Rd/ANN1JmbOn4nF04JOo6PcWg6+2JLt2SFpJZ70fxL8wVPnSXlQOQFeAQghOON9hqEvD1XG0VTvk2vpHoaGs6GdOal//PFHhzYHDhygvLwcs9nMuXPnmDBhAuPGjWu4waioXCXk8t5ykufKkcZFRUX4+fkRHx/PbRG3AThEvL3/7vuc/OwkYwaO4S+Rf6FrQFcsRyycLDvJ6szVuIe6U7qrFDpcOmfF9gosrSwck47Bc0ASIODCjgtIByU6j+hMl3u7YNQbiY6Ipnt8d1544QUkSVIjmZs4tXKAvPjii9x7772sX78ei8XiIN4RERG1ngU0mUwcO3ZMeZ2dnU379u2rtAkICMDb2xtvb2/uuusufv311yoG9j/+8Q/+8Y9/AHDzzTdjNptr1Yf6kJ6e3qjHb0qoY70ypKSkkJyczPHjxxFCoNVqqaioqHE/+wfAoTOHSNqaREL/BIZ0HcJQ36G2qI+LmanZChQDFTBz1kz+X/L/g/7geYsnZdYy9Fo9bQ1teTjsYTYe3kifdn34MPNDfNx9iDwfyScpn/Bk7JOc9T3LoIhBTHt+GqtWrWLMmDEsX76c4OBgsrOzq/xdV0d99qkr19I9LNMQGt2Q+gyqRjcW6lgbl4MHDzJjxgzi4+Ndnnvc2nGs+t8q3L9zp+fonuw6t4vBoYPp2LqjYpynpKTwz6f+ibAKDAEGRvx3BCuOrUDMvhidd7HSgMzMmTN5YfILaJ7TYBVWrFjR7tQitgqsN1jhAGgkDe5D3BHhAoFAr9NTWFZIN/9u7Hl6DykpKSQmJiKEYMaMGU32PrmW7mFoOBu6OU4iXkvOLnWsjUt2YTYdLnRQJuTkycx27dpx4sQJXnnlFdzd3encuTOn95xmuH44p/fYJu6G64fToV8HKvpU2CYCLwAlwE0XfyTgFuDRiyfzvfjbAKZ2Jmb2n2nbtuDivhf1W6PTEN4uHIDTe07TuXNn1q1bp/S5ud0P19I9XCsHyM6dO1m7di2SJFUR24CAAE6fPu1iT0f69u3LwYMHOXLkCNdddx3vv/8+K1eudGgzfPhwnn76aSoqKigrK+PHH39k4sSJtRyOikrzQPZce3p6UlZWxpgxY2q1nxw6J/9/eNfhZJ7KZNy6cbQZ0obTG0/TZkgbWt3ciiMRR3Db6UbZ1jLwgb5v9KVrQFdGdh/Jmr1rQMDI7iNJP5rO8hG2aA2Du0Ex4N+Kf8vh3KtWrcJisbBq1SruuusuB297bYmPj1fCy1UajobQaFWfVVRsocQ1hhFLoLtFx8jHRwKwa88ujJ5G5gyZo+gx/vBUwlMsn7+ciFsjWPOPNXh18OJ82XnbUsQbgaOgNWttx9RCj1E9OOF5gtPFp9GiZc5Lc9j+2HY+2PMBocZQpg2aRtLWJA6dOUSn1p3w9fDlmz++oZN/p9r33Q7ZES8v91FpHBrKhm6Ok4jXkrNLHWvdqYsGyROAgyIGVVnKl5KSwmvTXsOvix+nXztN74d68+bLbwK2JZK+mb7Me20eZ0vOYh1khU9xcECjx+YQkfG5+LvQ5pz+f//v/9m2PQceWg9KfypF+loi/lWbo7ylaOm1dA/XKgeIXq+nuLjY6XsnTpzA19fX6XuV0el0LFiwgHvvvZdu3brxt7/9jbCwMBYtWsSiRYsA6NatG0OGDKFnz57ccsstREdH06NHj1oOR0XlyuKqjnZN9bXlHBlvvvkmFRUVynIRV6TtS1PWLoLtQTBs1TBC54YyZMUQVvy2gtPdT8NzkNs9l8LzhWglLeV9ym1he95wougEH+z5gBe/ehGADYc3MGnzJBbtWMTTnz/NgHcGsOSXJUx4bYLTvo8ZMwatVsuYMWNITk6moKAAb2/vWot9S3lANEUaQqNVfVZpKcgJ8OqaFLUm5v84n7ZvtOX016fRzdWRvzKfL5/+kttP3U7c7XHM/3E+Ny++mfd+e48Vy1bwzrx3mP36bHb+sJPivGLO/3IeSmzLDr2OevHYvx6j5309eeGZF9Bfp+eGu2/g1btepY1XG7zcvVi5bCVr/rGGiOMRfPS3jwDb8hvzDWaWj1hObL9Yegb25MnwJ4Gqz52aPof6JqtWqRsNZUPbO6nLysp4//33GTZsmEOb4cOH880331BRUUFxcTE//vgj3bp1u+wxqKg0BnXRoOryGMXExDBmyRgO7DxAXk4eW97ZouT/WPbrMl6f8jr5+flYS6y2HW6y21kPDMbm4Oh58fdd2CKp9diiPXTYli7ugOBWwXj386Z9fHvC7w93OY66PodSUlLw8/OjdevWLr87qDQgtSkjExkZKe68805RUVEhKioqhCRJYufOnUIIIQYPHizGjBlzeXVqLhO1xGLD0RLGWtuyiA0x1sD2gQIQXv5eIiMnQ9leU0kquY8JMxNE1NooEbUmymF/+zrhQggROjdUkIhwS3ITkzZNEm1mtKlS6rZKjfJKZXDl9roknYhcGSnCF4UL81Kz0E/WC79pfkJKlESbGW2UMVVXTquhao43Bi3hHpaprbapGr2lUY/flGhuY72cMrX1HWt9S75W1t3KGEYabKUP9TiUw5VL3raZ0camuT1t7wNCq9WKB0Y9IAwBBnHjXTcKfBA6vc72vmQ7VvCjwWLWqllC6ikJSSOJB0Y9IGI3xNpK3l48vhC28uhSoiTCF4U7HWdlja3pc7ica3M5NLd7uDpqo20Nqc+fffaZ6Ny5swgNDRVTpkwRQgixcOFCsXDhQqXNjBkzRLdu3URYWJiYM2dOg4yhvrSka10T6ljrTl01qLr2ctlx/yB/MfDZgSIjJ0Nk5GSI8EXhwnOEp5A0kk23fTUieHawUtrWc4Sn0CZpndvQPoiZM2cqJXLx4VJJ3EgULXbWr7o+h2T9vhJ2sita2j1cnbbVygGya9cu4e3tLcLCwkRCQoLQaDRiwoQJwmw2i1atWol9+/Y1WGfrg2pcNxwtYay1/aLdEGMd+OxAgQ9C6iUJQ4BBEb+aRF3uoyHAIPRT9EI3TCe8/L0U0Y5aGyX0U/Qiam2UEEKIkLkhiiB7TPZQ/i8lSq6dH3YiPXPlTGW731Q/EbUmShinGUXI3BARMidE9P1vX9Ftfjexfu/6RjOKr5Sx3RLuYZnaapuq0Vsa9fhNieY21stxfNZ3rJUdGTU5NmRkgzVqbZSI3RAr1u9d77Cfsa1RMVDRI7oO6CoMAQbhPdJbxG6IFZM2TbLps3TJAaL86C46RnwQUk/JoY2nwVMsXbdU2abVaoUQQiTMTBCGAIMYnzRexG6IFfO2zRPhi8LF+r3rnY5r7NixQqvVirFjx9Zp3Fea5nYPV0dttO1a1ueWdK1rQh1r41Of54msgwkzExQbtNv8bo7OardL+kzkRfs6UhLoETNnzRS63jqbdkdKQvK1OVIkX0nM2zbP4RzOJjIrP0dcMXbsWCFJkvDy8rriTmmZlnYPV6dttVoC06tXL77++msCAwN57bXXEEIoSZy2bt1Kly5dLjMORUWl4ahLCdaalqrUxJsvv0nU0ig8sz0pyi1Swt9qKkkl9zHqmShCjCHovtVRnFeshO0hwCqsbMnaQvf/687DYQ9jcDMAUGopVY4j7BcxXiyxqCDXLd8KGjT4uPvgpnFjZPeRFJQWIAmJ44XH+ePsH+w+vZt7O93L8K7DHfrekKHkapmuxkPVaJWmSn1LYteXlJQUhvQZQvdj3ZVQ6erK5MoalzgrkbcffpvCpEIOfnmQFbtXkLQ1yWG/CS9MUDRW76Xno/c/YlvmNmKiYzB3MLNi2QpK3yiF9iBpJP7yl79cOlEF8BtQCOKoQLrpkliXlZVx7tg5/nLLX5RlhgCJzydy7vQ5PG7xYMXuFWTlZ7HznzsZ3nU4UDUkPD09HYvFoiSxcxUyfrnPPZW6oeqzikrDcDnPk1PnT3Eq/xT/7z//j17He6GTdLD74pvl2P5/0WYGEFuFLS+IBmYvmk34jHDcbnGzLY/xseVvysrPIm1fGgPeGcDC/y7klu63MGjCIDJPZSr6m3403enzp7IOp6enI4SgdevWqp18BaiVAwRsmao3b97MuXPnyM7OprCwkC1bthAeHt6Y/VNRqTN1+aJ9uWugw9qG8e6D7zJ36twqolydkSn30eMWD04Xn+b2sbcT2D6QAY8NIDoimtAjoYhZguyt2ezN3cvGwxu5uf3NrjtS2fkBtvWLPqC/W4+12ErhtELKt5eTtj+NTVmbOF9+no6tO3Jf5/sY3X20Q7kwmeq+ODhDNayvHqpGqzRFGtPx6cxBa6/p8vvmDmbG3jS2Wo2bNW0WhWcLKTtfxpGPjzD2prEk9E9w2O9s2Fm8H/TGEGAg7qU4RRdlIzf7k2woBP0FPbtP7GbJx0sIvTPU4XySRsJ/iD+a3y+aXxJYrBbKysrYmbGTX//8lRdnvegwLnMHM8E+wYT6hVbrkK7tlwM198eVR9VnFZXLR36eADXamrL+T/9uus2OnZNK2fkyCs8Wsua/a7AK66VcIG7Y/i9XUcT2W/KV0LTSkJWfxfIRy3ko7CEeePgBIhdFMuaxMURHRJO0NYnc4lzKt5Q7TmReJDoi2unzp7IOX+nJgmudWlWBsUev11fJOq2i0lxpqKokzrLvy+I2fvx4pU1lZEGMfiyasDcvzdINmTuE8oJy3Da7Yf3GyqmwU/ya8avN89zn0v4SEh7CA72bnnLKOV9xHl93X8KDwml1YyuMTxgZ2W0kWV9kQSHovtHx6suvsv3P7SAg7o44pwmlqvTPyRcHGfvkpvaCrnqwrw6qRqtcK8jOC0CpjmWv6c7er4ysbaXPlPL2G2/jrnVncuJkYobY9Mu+0lbBhQIefPRBDD0NvD7ldTRmDV8M+oKb291sq64VuYbsT7O5ccSNRK2LQmfR8cfAP2jl14pzX54DwO1eN4puKkIqkZC2Soi7BGyy9aXcUk7Uuih6tO3BhkMbAJvzI+aTGEoqSli6aynHCo+5HE9tq8Co1biuHqo+q6hcPrWxNeUEqEHeQdxqupXfR/7O3vf3IiEh9ZfQu+kpfbCU8gfL0Uk6/t9t/4+NhzeSuTGT8jnl3DTqJpK3JXPuwDnuibjHNuE5omrJ6oT+CSRtTaLfc/1Y+9+19BjVw8Fmtq/eaE9lHa5rFS+Vy6NWDpCaZgkkSeLVV19tkA6ptFyaYhUQuR/2S1fqi1yiSy4jGx8fz/jx47FYLIpIV24DcPjMYUavHs3UgVPp1LoTkzZP4sJtF/DY5IG1zIqlwMLx74/bojy2graPFgsW20kFeLl5cabiDP56f36M+VE5rnyuTq07UX5dueJZjvlL7cfoSrjtsX8QqYb11UHVaJVrgcrPEGcOWnsjMnFWIoVJhbyte5vub3TntuG3VdFfmV9O/EK5tpw7HrtD2f9fyf8idU4qIcNDOBx6GHaAx3celF4oxVJsgf/B/vD9HM4/jL+XPxtmbyB1bCpvx79N4duFSDdJaB/UcueoO9l882ZKraWUUUY3v27c/NjN9H2pL0t3LaVLWBc0Og1BI4I4UnCEHm16KDOGUeuiyC3OJcArgIT+CaQfTa/WIV0bVEP7yqLqs4pK/ZF132w2k56eTnx8vEtbM21fGpM2T0KLlj8K/8DH3Yff//c7B7ccxHK9Ba1Oi+kBE8dvPM658nO0oQ35Uj4VooIfsn/A5GPity2/Ic4KDqw7wJKBS3jQ88Fq+ze863BOfXOKF+a8gCRJPNTjoWonFmVUHb7K1CaJiCRJLn80Go3QaDQNkqykvqgJ9hqOxhzrlaoCUlvksTZUv5xlfK6c+LNym9gNsUKbaMs+Hb4o3Pb6YjZqKVESumE6oTVqqyRnclXtxf7c9udqzOt6OclNGyMxakv6e62ttqkavaVRj9+UuJbHWletrpxV35lGy9vkiiuB7QOV99yMbkqyO4/JHkJn1CkJUDW+GiENk4QmUSNC54SKqLVRSrI7+ZyAePrTp0VGToaYt22e8JvqJ8xLzU6T4W3ZssVp8tbIlZGi6/yuSuLTlkBLuodro23Xsj63pGtdE+pYGwdZx7VarfLblc0YvijcwSb2SPYQHn4el6puyZVcKhUKIPJidS+9XTLUi79nps4UwY8Gi8D2gTUWNmhK33HqQ0u7hy87CarVaq3yk5eXx7Jly+jRoweHDh1qYLeMSkukqa5va6h+OVvnV3n9u32bzFOZHM4/TDtDO0KMIST0T8DcwUyrjFZIsyWuP3Q9Qx8aivFFIzwIPIfD8pfKOT983X1ZtXsVvlN9mf/jfJfrDutKTXk9LmeNv7oWvWFQNVrlWqAmra6sVfHx8RiNRvz8/IiPj3eqifK2/0z6DyaTicmJky+9NzEaNz83Hhr/EONvHs/Lr7xMYPtABv5jIHfcdQfiE4HhMwO6XTqWP76cp5KfYtkvy5TsapJGYtnjy1i9fDXP/OUZzrx4hkfKH2FInyFKH/+V/C/c/NzY8/sepT+pO1OVCL5v/vgGTzdPOrXu1AifqMqVQNVnlWuVhsgLJ+v+mDFj0Gq1SlQ1oCxPHLd2HGn70jD5mAgxhtAjoAceGg+0Gi1ud7vZcntUzvFhVyiArdgSnpYAR4HnQDoo2d4vhOxPs8n5M4dXE191moep8rNGpRlwud6V2bNni6FDh17uYS4LdXax4VDH2vjIs3xRa6KE+2R3oU3Uiqg1tnK3sRtibZ5nEG5ebrYZyMiqkR7OfuRIErnUrT2XM9bGjNxRI0CqpyG0TdXoloU6VtfYa1V1ZQkrR2DUVocWL14s/Pz8hNFoFJIkKbOKWl/bzKSb0U0EJQQJN083hygQSSMpx66sp3KUycw5M0XshliHKJWMnAwRvihcGKcZHaJWmjst6R6+XG1r6frckq51TahjrUpD2I/2+uwsqlq2o7vN7yaM04wifFG4yMjJUPR+3rZ5ImRuiPCf5i+8Er2ElGiL3DOONgqtr1bohukcIkCkSElEvhcpfHx9BCDmzJkjBj47UAS2DxQDnx1YJYqwJdHS7uHLjgCpDrm8l4qKSs3I5bLe3vU2BaUF6LV6rFj57o/vyDyVSahfKBqzBnyg3FpOeUG5LUHebGCH82Nq0eLj7sMT4U/grfMGwMvdi8xTmUoW7JKKknr3uTEjd9TSuI2PqtEqTZHGqBhlr1XOKljJ26Z/N91hFs8+Eq1yvzJPZTJowiCCrgsiNjaW/Px8CgoK0Ov1tig8LQT1CMLTz5O2d7TlvPY8d0TfQWD7QELuCgENCKtwmek/emI0OqMOTz9PoiOiiY6IplNWJxaPXczq5atZPmI5j/d6/LIj+VSaJqo+q7RkGsJ+lPX5n0/9k+OFx9m4YyNv73wb7wBvSn8qpaNfR7QaLae/Oc25aef47fPfePrzpxm9ejSHzxwGoLCkkPySfDQ7NIjZAnZC50GduXPunfjf4Y+EBO7AYHjw0Qcx6o10H9Md/yB/vPy9ePPlNzl5/CRvvvymEkWYkpJC0HVB3Dr+VsatG0firESnzzS1OmIT5XK9KxMnThQdOnS43MNcFursYsOhjrVxkdcntpnRRkSttUWASImScEt2E+GLwkW3Bd0uRXXI6xP1ldYtXvyREiXRZkYbZW14Rk6GiFoTJULnhArtMK3Qt9aL4EeDhXGaUSxdt9TpzGdLpCXdww2hbapGtyxaylhrMzN4OWOtLgIkak2Uwyye/axi5X7FbogVGl+NLZLjYtSHJNkiOgLbBwpA+Pj7iA4PdBCt/FqJu5+526bFa6OEfope9HuqnzJzKOcIcZUDRMbDYFuz7mHwqPf4mzIt5R4W4vK1raXrc0u61jWhjrV+VNbqylEeixcvFpLGpr2GAINDpLSXv5eIWhMlQuaGKNvwQbgluzna0nq7n8r2dKRjfhD3ZHehn6wXuuE64WZ0E7OXzBYDnx0oTCaTGDt2bJVnBT4I/RS9MAQYnD7Tmlr+w+poafdwddpWqyowf//736tsKysrIyMjg927d5OUlHT5nhgVlSaGs4otld8zdzArWfldZX1O25dG0tYkEvon8ETvJziafpSwgDBGdhsJAo4VHuPXnF85dOYQ5hvMFJYUcubCGcpvLqeiT4Ut8mMrl9YtAho0tHJvhbvWnde+fo1JX00i0BDItuPbaGdohyXdgqXQQvan2fS+szeA03KQ1Y1RpfmgarRKU6Aulb4as2KUfT++P/Y9Q5KHKH2aM2QOmacy8ffyV6IqKmfjj4uLo6ioiJSUFKKHR7P7sd1kfJTB4LsHK1UIYmJiEELw8ssvY+hn4Pevf0cUCn5I/YFeb/Xi+tuvh8HQeXBn+g3vR0pqCl/f+jXuA90hGnLfy2XVqlWMGTOG5cuXO/TfXetOKaW4a90b5DNQI+yuLqo+q6i4pnK5cmclblu1akVJRQnPv/g8oyNGs+2Rbfy2+jdMD5hYsWwF0tcSdMCWv6MDlM8st9nMfbBFUctB0Hou5QFZC+zGVg9VYIvo6w+9Anuh/UXLto+32bafg4yPMsj5M4dVq1YpOUji4+N5NfFVQoaF0Ll7Z0JfDGXJ3CVVnmlqdcQmSm08KDfccIPo0KGDw0/Xrl3FvffeK5YuXSqsVmtDOWvqhTq72HCoY72EvBbbmedXfi98UXiN6wHDF4ULKVFSqrzop+iFfrLeoRKM/bpFuX1t8n7Y//hN9RP6yXoR+V6kCH40WEi+kvC7xU8Etg8UK99f6XTm0VlVhMamMfJ+2NOS7uHaapuq0Vsa9fhNiaY81oae6arvWO37UblPrvKAONtf46kR+tZ6kTAzocp+VqtVfP755yIhIUEs+HCB6PdUP+HmbZf7Q0L4TfMT87bNExk5GUqVGZ1RJ+Ztm6fMaEoaSWTkZIgNmzYox5c1MmFmQr0j95ryrGNTvofrSm207VrW55Z0rWtCHWv9qCkCpDot6za/26XID72LqGndpXxMDjn1JLvtPgjPPp7CEGAQCTMTHCrPrF69WumT/feAlkhLu4cvOwLk6NGjDet1UVFpBsizg6vmrqri+V2waQFbf9/KE72fICs/C3MHM+PWjqOgtACj3kjc7XFKREVC/wQmfTUJk48Jcwczh/MPczDvIOYOZuU8eRfybDIMDOk4hF9P/opAOEZ/9HHsn7/en3JrOdf7XI9FWBjfdzxZ+Vn4Zvry05afuPuxuxWvde6pXOY8NIfKyGO8kuvLnXn3VS4PVaNVmgJms5lVq1ZhNpuvaj8qz7jZ/7/ybKOr/f8d92/KL5RTcqaE5BeS+faPb/nV9CuHzxxm1+e7KPi8gP639+eBJx5gn9c+Og/uzI8rf7x0kJsgvySfudvmkpWfRc/RPdm2chviLsFLb7yE0AqwglsvN1J3ptKnrA8r9l/sV8wcYmJimLhxYo19re1noHL1UPVZ5UpRn6je+kaLNUQEsf25w4bYjlE5Iu/J2Cd5Y+obmB4wkXkqUzlX5qlMAg2BHAw9SMWuiktVXMAW6SFHTeuAiovbLtrRPfx7UGGu4MDWA9xx/x2cu/8cfyT8QV5unhLFkZycjNls5sSJE+j1eo4dO1avMao0Ua6gI6bRUGcXGw51rFVx5vmtHLURuyFW6CfrhTZJK/RT9FUiKuT23RZ0E6FzQoUuWSe6ze8m1u9dLyJXRgq/aX7C+zVv0W1BN2GYYrjkofap5MlORATOCBTdFnQTke9FOo3esPeWy31fvXp1vT6jxojWUCNAak9ja9uVQtXohqMpj7U+UQfV6UFjjLU2ESBCCLF+73oR9HCQMkvoH+QvYjfEiq7zuyq6rPfWC3dvdyF5SqLfU/2EYaTBNpM4wlNcN/M6QSRCa9QK75HeInJlpOi2oJsImRsiJF/J4ZiVI0Dq2tfmRlO+h+tKS9BoNQKkYWgKY61PVG99dHvLli0NEkFcm3PLkdO6YTphCDAoz4qotVFCm6S9ZCfjGNFhn+PD299bGEcblW0hc0OEEJc0dv3e9aLrgK5Co9GIsWPHKu8ZAgxi5syZSv8a23692jSFe7ghqVcESF2zUt911131csCoqDR1KnujweZ53vr7Vo4UHFE84HnFeUoESOWIiuiIaLb+vpU9p/dQbinHipWsgiyStibxW85vWIQFvVbPwbyDVIgKwJbnw9rfiuZrDZr+Giqwbc+7kMeZkjOUVJRwq+lWDp85zLBVwzhXco6c4hwiYyL5JOUTxZsfExNDenp6vcbeGNEazj5PlbqjarRKU6M+UQdXOiIsrG1YraIphncdzvBVwxnUdhBb3tlC77/1Zubgmbgdd2P/XfsRXwvKysuwXrACsGfVHjz0HvhF+vHAjQ+wdvZa3M+6U3a+DGmrhPERI6eLTzO041BCHwtl+zvbsZRY6H6sO2FDwjitO82cQY79qm1fQc350dRQ9fnaIzc3l+Dg4Kv6N1ifqN76RotVd67MU5lMeG0CGR9l8OA/H2Rb4DYS+icwvOvwas8t69iTsU/yxaEv2PXhLm4ccSODBw5m0+xNFJ0pYuKkiTyb8yyd/DoBIPWXEJsuhlDfiC0PSP+LNjRW6AMX+lzgPOeV8/Zo2wOASV9N4tP3P2X+1/OxlFrACp999hnBwcF0ebALor9Ao9M4RBPaP69U3W3GuPKMSJIkNBpNjT9yu6uJOrvYcKhjrT21nZ2z9zBHrY0Ske9FCvNSswiZGyLMb5tF3//2FX7T/MSYVWOELlEnSMQWTZKorTbnh3uyuwhfFK5EnsjbwxeFN9hYq6u/3lRpSfdwddqmavQlWtI1r4mWNtYrHQFS177I+r3rz11i5cqV4oEHHhC+/r4i+NFgMT5pvDAajcLLy0uZeQxsH6jMavr5+SkVYOZtmyfCF4UrVbvkCjKB7QMbZKxNOedHZVrSPexK21R9ttGSrnVNzJs3r95/g83FvpKpTQ49OdJNZ9QpefBcIeusrIuGAINDBHTshlgx8NmBQuOrEVKk5BDdoVR3ibTb5mP7bZhiEB7JHpfs6YvvJcxMEEIIW+XFi+fxMHiIwPaBwmg0KtocuyFWbNi0QelnXfKTNEda2t9rvSJAtmzZ0ti+FxWVJkfmqUymfzsdJJQ8HvbrHAGHNY/OZufkqi9P9H6C7ce3s+XIFo4XHSdtXxrjLONYNHsRFXdUkNc9jz/O/oFG0lBuLef9/e8jdgjYCqX9S2l7Z1tyinOU43pqPSmzlGHBgqmVifCgcJDA5GMCUCJAEvonNNjnYR+tERwcrObuaEKoGq1yJXE209UQs1/VRYQ19Ixq5qlMpn83HQTE3RFXZd26PLv3z6f+yfHC4yQ+n8jq5atJnZbKxtCNZO3LQmfRUXy+mFZbW/HWirf491P/pk/HSwmaeozqwUM9HlI+l08Mn/D5wc/J2pbFubJzpB9NZ3jX4fQY1YPT75ymx6gepKSkUFJSwsGDB+s9TjXnR9NC1edrj3bt2mEymer1N9jUc6PVVeujI6JZ88Aasj/NpvvI7miDtNXapnJupl6jeuG21s0hAqT333rb7O8ISB2aSlFZEUt+WWLLk7eVS3k/tmLL8bEVKLT9Lrm5hE6tOxFkCGJf7j5Obj0JhbBk7hISn09k6sCpPBP5DMWbi5maPJXbht/GIy89QtGaIrr06cJHMR/RM7mnw3PIfvyq7jZjrqAjptFQZxcbjmt9rHIuD/s8HvI6x8iVkaLNjDai1eutlPcycjJE1NooEbUmSokE6Tq/q0NVFrmiiyZRo9QJl9cnahI1jpEddh5v7yneyna3ZDcROidUeW1earZVevGRRMDfAkTI3BARuTLSZTSKs/XldaW5zFC0pHu4JawvF0LV6Ibkao3V2UxXY89+1WZGtS665KwKV+VjydVZDAEGIYRwrOwCwt3bXRgCDOKBUQ8Ik8kk+j3VT6k64ObtJqLWRjlUcgl6OEiQiAidG+qgwfYRhCaTyWGdeW1oLnrsjJb099oSNFqNAGkYLmesTf3vubLW12asssbJVVWcjc0+QtqVjeos2rrbgm61igAJnB7oUIXRe6S3Qx6RysRuiFUiqt2MNu2fOXNmi4ryqI6W9vdanbZprorXRUWliRIdEc3o7qMZ3X20w/rGsTeN5WDeQXKLcxE7BKueXEVKSgqpO1NZvWc1q/esJnVnKgCd/TujlbT0CurF6O6j6X9Df0KNodzf+X6ef/F5vPy90Jq1+Lj7ENY6DD165fxSf0mpUW5wN1zajkRYYBh+ej/cNG6cLDpJ9qfZiEJB7sZcjhQc4fODnyt9sCclJYW9mXtJSU1x+n5tiYmJ4dixY01ydkJFRaVxiY+PrzK76WxbQ1KbGVX7mdO0fWlE/DeCtH1pTttW1veUlBSCg4NJSUkh81Qme4L38FTCU3j5e9FjVA92HtuJxWKx7awDNz83uj7cFf0LetLT08nOzubn939Ge48WfWs9QX8N4pMDn5C6M1XpV3l6OeFB4cy+dzZzhsxRok7kCMKwtmHEx8fj7u5e79wpKioqzZua7Ct7rbpSyOdMnJVIlwe7ENg+sE4aJWvckrlLyM7OZvz48aSkpDiMZdJXk5j/43zm/jAXgENnDjFu7TgGLBtA8KxggmcHE7kqkrd3vc3oD0czYNkAQueGopW0eOu80fTRwKCLJ9wATAOj3kjkwkgyUjP4e8Tf8dB6EGoMJW1/GoF3BPL6p6+zJ3gPafvSmLhxIpmnMsk8lcnEjRMxdzBzX+f76OrfleiJ0QS2D0TfSo+Prw9FRUVERUU5XIfK10U+TuapzAa8EiqNgSSEELVpmJGRwZIlS9i/fz8lJSUO70mSxObNmxulg7Xh5ptv5ueff26046enp1/1sn5XiqY61sZINFTXscpLW/5I+IO8nDxMJhMbd2xUQqpHdh9J+tF0zB3MzF84n59W/oToL3jmqWfIPpethF2DLdzv12O/svXPrbYkTReR5H+ShKfWE6OnkXaGdhSVFTF10FQ6te5E6s5U5Ry7PtzFdfdfx7mbztGjbQ+mDpxaJaw7ODiY2NhYEqclsi1zW73LlTUXmuo9XB/qom2qRpsb7fhNCXWsjqSkpPBq4qv0GNWDPzv/yb68ffQO6s3Of+5U2tgn5JucOJnbht/GhNcmsGXBFqxWKyaTiQFvDmD1ntUMDh1M9rlssvKzuFl7M/u/2U/21mz8h/hz/qbzDO44mI5+HfHN9OWNqW9QenspNw6+kZuvu5mPln9ExZYK7h10L99+9S2SJPHsy89yNuxsjaUiK4+1pmdec06+15Lu4dpq27Wqzy3pWtdEQ43V2d+2vATZZDJdsXKs8jkNAQb0L+gZe9NYZdl3bcYqLx/3zfRlStwULBYLhgAD3u7e5PyZg8lkotWLrdibuxcPrQeebp6EGEPYe3ovJZaLfyM7gK2gC9VRkVVhK23bx+4kApiDbcmLjARBDwcxeNRg0vanUVhWiIfGg1JrKVq09AzqybHCYwT7BHOs8BhjbxoLwIrdKxjacSj+Xv6KXk/cOJEOZzvwytOvUJRbhFarxWKxKNfBz8+PgoICjEYj+fn5Suly+8+qOdHS/l6r07ZaRYD8+OOP3HzzzWzYsIEvvviC/Px8srKySE9P59ChQ9TSh6KiUm8acrZL9tjm5uY6bLefPbT34sr/79S6Ezv/uZOn//M0hgADT8Y+SVjbMN4d8S7vPvgua/as4a3tbzHpq0lkfJTBudxzFG0qYup3U/kg4wPez3yfO5feyaTNk+io7cjOP3c6OD/YAWK2wLrDilVYKaooIvtcNntO72F/3n6W7FyieNSHdx3Omy+/SdTSKFa+vpKsCVl8POZjpwa2PLs4+/XZLd75ca2iarRKU0DW1sqzZI1F4qxEnnvpOUKGhbCj/Q5KK0rpGtC1ylrz1J2pbF68mZw/c3hqwlNM+moSW96xOT+0Wi3x8fEUlBZQbi0n83QmR/KPoC/Xk1ucy6TYSUQti8LnNh8EAqPeyJwhc0h8PpGRi0fi1teNm9vfTNztcVi+tFBeUM5naz6joKAAb29vzoadZcXuFUx4bUKdPhP5mffCCy843U+NyGs+qPqsUhcq27spKSkUFRXh5+d3RXNNyBF+z7/4PGNvGlunyjJwKa/H2bCzLFy4EC9/L0puKyFwaCBe/l60vrc1gd6B+On9sGy3cDbpLHue34NhtwEtWttBLubzqPi1wubk+BSbUwRszg8Jm1NED+gubT+94TSrMlZRWGbzjPQM7EmIMYQbfG/gid5PMPamsST0T1DGJUd6I9kcIXK0dHRENJ5lnmjKNRiNRsaMGeMQlShJksNv+Th1/axUrjy1coC89NJLPPjgg2RmZiKEYMmSJRw9epSvvvoKi8XCK6+80tj9VLnGachQa/nhcuLECYftSVuT2HVyF5M2TyJqXRTLfl1G6s5UUnemsuzXZUStiyLzVCZnw86if0HP2bCzjgeWwCIsHM4/jPsAd3RGnU2YsYmjp86T/JJ8PjvwGRO/n8hZzl7cTcJNcnNI3CSwGUQaNJyvOI8VKz8d/4luC7oxbNUwJVnrop8X2ZK2ViLzVCbj1o3j1vG38mriq7Rr1041lFswqkarNAVkbV21alWtHNaXG9Y9a9osinKL+G31b4QYQzhTeoZ7O95bpdRidEQ0HjoPAKzCyqH8Qwx4bACB7QNZuHAhAJuf2Yy0U6KLXxfa/tSWMzPPcODXA6xNW8u6f6zj5NaTtG/Vnu3HtytLbOLuiOOpm59SEqrKui2EwGg0Eh8frxjEGR9lOHwmNYVKy888SZLIzs4mLi6u2tBrlaaLqs8qdaGyvZucnKw4VK+kHSc7WROfT3RYvlcdsq6l7UsjrziPoZ2GEh0RTUxMDD1H96Tiywp2r9xN2R1lZARn8M0f33BH8B1IX0uIC4LS86XkbszFihUJCToAEhhDjTZnh8BmKwtgJzD74olfBF6B1n9pbbPFr7eglWxOFE+tJy/f9TJ3BN/Bn0V/sv34dmUisfux7gzpM4Tv075nzpA5xN0e5+DACGsbRmlBKYVnCzEYDFXKVU+fPh2TycT06dOV9rX9rFSuLrVygPz222+MHTtW8XDJa2LvvvtuXnnlFSZNmtR4PVRRoWFnu+SHS7t27Ry2J/RPoHdQbzr5deJIwRFCjCGYO5jJK86jnXc7jhQcUSrADO04lMP5hxm3bpyyjnBkt5Hc1+k+9Do9f3b+E4//eEAfm4PDKqxc73s9WrRYsVJBhXJegUCj0dicJRfzf8h08e/Cda2uQ4OGCxUX2Je3j88Pfk7kqkjS9qdhFVbbQ6EScm6SbSu3kfNnThVnj0rLQtVolaaArK2VZ8lc4Sqyz1WUXmWef/F5DAEG/jPpPywfsZzHez3udOYtrG0Y82fPxz/In+CRwZjzzOxfu5/JiZOJiYkhOTmZ4rxiPL71oFdeL05uPkl5YTklaSX89M5PFOUWYfnSwrHEY+z7ch9JW5OU49obu32i+ih6bDAYiImJ4fu07/ko5iMG3z3Y4TORZ0dd5WWSn3mygS2EcIgIiYuLU3OANBNUfVapC5Xt3cbMteTKkVrZQevKYVt5u6xrL371Iqv3rAaBoo+ZqzKhBMQFgfu37nTw7YBOowMJ+jzcB3dvdzRuGigDsUPQWt8ajgICCk4XwE3Y9PWGi7/tJg116NCg4czeMzbnyFEos5YBcOGnC4y9cywHNx207WdnM1d+BoW1DVOcIvJnYp+LqnJ7NRKv+eKyDK495eXleHt7o9FoaN26tcOXqS5dupCRkdFoHVRRaWjkMlbp6ekO24d3Hc7wrsMdSuGu2buGDYc3MLTTUPw9/TF3MCsG66bDmxAINh7aSElFCV8c+oISSwnnys4hEJwvP68c2yqs/J7/OxZhcRBfD40HHjoPisqKbOsa7dY2+uv9mTpoKmv2rAEJ+rbvy1vb3+L0N6c5svEI9Ic2d7Yh7va4KmOMjogm70IeBx85yJGPj1Rx9qi0LFSNVrna1CcnhasSgq6i9CqT+Hwiic8nKq/lNdf2pctl49u+fGHQdUHk/JnDP/75D44XHic+Pp7ExERuv/12Si6U0PPhnnyf+j1WqxWdRofJZOL8+fPk5+fj9q0bCTOcl3NMTUplgrst14j97G12djbp6ekOa/ftk2y7wv4zlY91/vx5srOzMRqNjZqAVqXhUPVZ5XKorlT45eKq/K7syACbrlZ+7aqdrGeH8w9z9OxRxd5NSUmhqLAIsEVEz506l9uG38b076az/fh2jl9/nLBZYRyJP0LBqQL03+nxvduXvP55SFslWg9pTd7GPJtz4/eLJ+8PbALKoGJthc1Z0gHb7/4XI6l3AJ9CkbBFCo5OGa3YzCkpKZw/f16J1nP1mQQEBDhot1r2tmVQqwiQjh07cvz4cQB69uzJ22+/jdVqxWq1snTpUoKCghq1kypNn6YejltTuLH9+2Ftw/D38mfDoQ0gYOxNY4m7PY45Q+aQfjTdJvYSjO4+mlC/UEoqStBJOvbn7edowdFLy1ckDSaDiRBjCP38+3F7+e20cWtDW6+2BHoFEuIbwgejP+D7J7/nBuMNSEiEGkMJ9Q0FwNfDl6StSazfv56MUxncHXI3e5/ei+5bHRSC7hsdKZEphLUNq/L5y7lJflj4AyePnyQgIODKfNAqVwVVo1WuNrXN02SvVfazZ/Ya7CpKrzrsczhN2jyJ+T/NZ9Jmx5l1+Rwhw0KUcOpZ02bx17/+leeee46bb76Z813Ps6/TPgY8PUAJbZYjMYxGIwYMnPrmlNM+hLUN46s3v+Lk8ZPcNvw2Jm6cSGRMpJIzqnLbmkKlKxvi9hEhM2bMUGcemwmqPqs0VVxFl8hL93wzfQkODsY309dpbovKOS9kXZs6cCpP9XmKuNvjyDyVyXMvPYcQAq1Wy3//+19iYmIU7Tu46SDnpp1j96LdFJ4rBD2U3l5Kdnq2LcqjPxR0L4C7cIiS1vTRgDtQAuzGFg1yFHjuYudmY3OQXMwVIpkl/D39lfMmJyeTn5+vROvV9JmAGvHRkqiVAyQyMlKZLX/ppZfYsGEDPj4++Pn5sXLlSp577rnqD6DS4mnqJflqCjeu/L65g5lgn2BGdh/pYKTKYj+y20iQINA7kACvANy0bgC0b9WeyBsjMV9vpo1nG3LO53Ds7DH2nd7HacNp/hLyF/y9/Mkvzed40XGStiZx6Mwh7rj+DsbeNJaPx3zM7CGzCQ8KJywwjCMFR9Dr9MryG4DJiZMxmUy8NeMtZa17U//8VRoXVaNVrjb2RmN1DnFXWmWvwbKRWRfHbdLWJH757BdG3zqa7R9vxyIsHMw76PQcnQd3pt8/++Hl78U/n/4ny5YtQwjB448/zoQ7JjD2prG8+fKbDoZuTEwMkiSRn59PXJxj1J2z8crn2ha4zXnOqFrgzBBXDfDmh6rPKk0VV3pSuYTtkrlLnDpsKztyZS2Uc2qEtQ0jdWcqor/AEGBg4cKFjucSYEm32BKd7qrAesEK7iD6CMq2lEEhiK3iUvR0GbAJ/Pf4c/+N9xN0X5AtAaoWNJ4awh8Kx0PjgfS1ZJso1OgwBBgYnziemOgYBweOK0eHqrHXBi4dIF9//bXy/8TERBYvXgzAoEGD2LZtG7GxsTz55JNs2LCBf//7343fU5UmTWOuUbwc7Gt7V5eZWc7rkXchj8xTmaQfTedY4THSj6Y7tAtrG0Z0RDRJW5P4MPNDvvnjG/4o+IPc4lwkSeLukLv5eMzH9G7Xm5ziHMpFOeXWcvKkPHac38FnBz8jKz+Ljn4daevZll9P/spzG5+zRZtIMP276SzZuYQebXvwZPiT9PmzD9ZZVvr82Ufpe0xMDPHx8cTFxeHn50dKSkqT/fxVGg9Vo1WaEvZGo+zkqJy4E2qecaxr9nxZ45/o/QRu37hRXlBOeXo53QK60bl1Z4eoP/kccXfE8cPCHzj420Gs7lY+KP2An0w/cZrThLUNwzfTl1u638Kt428lbV8a49aOY9y6cVRYbbmbCgoKlDHJs5vZ2dlMip+kRKHI55IrDZg7mKuNQqzpM1VpXqj63Hy5GhHNV+KcriKh5ei5+T/Or/K+2WxGq9USGh6qaBtASUUJ49aNY9jKYYxbN07ZR9b+VxNfVY5l7mDmxsE3suKbFVW0bGT3kbZlKzJyVRdQEqA65Pwosf2c2XiG7//4HhEh0HpooQI88OC3t39j8G+D+cuYv6BvrefmcTezLXMbb8W/VcVRk5yczJOxT7IneE+tdLmmSPK6tmvqkfMtHuECSZJESEiISExMFFlZWa6aNQn69OnTqMffsmVLox6/KdHSxhq7IVYEzAgQsRtiq7xXeaz2bTNyMpTf9mTkZIjwReGi1eutRLf53YR5qVmEzAkR5qVmEbU2Smm/++Ru0WNmD6FJ0AgScfjp+9++ImpNlDDNMgkSEd6veYuotVEiak2U0E/RC22iVuin6EXshlhhMpkEIALbBzr0R94OCJPJVOPncLWv6+LFi4XJZBKLFy9u9HNd7bE2JNVpm6rRl2hJ17wmmsNY5b93o9FYa41yRm3Gaq/b8nnHjh0rDAEG4T3S26n2CyHEH3/8IV5//XVx1+t3Cf1kvdBP1ittDQEGm75KiOBHg23vT9GLgc8OFBqNRgDCy99LeU54j/QWhgCDCH40WEiJkghfFO5wLvm5YZxmFFFro0TshliRMDPBQRObw3VtKFrSWF1pm6rPNprjtZbtq7rq1uWMtb7nrAuu7OHwReFCSpREmxltqrwv98vNz81B25anLa9irwpxSfsHPjtQOZa9Rla2AaPWRAl8UPSWSJudrE3UXtruc9F+jkSgv/gTecmmliIlYQgwCEmSBCAkjSQCZgSI8EXhLu1/eVyGAIPLNjLyda3u+0RtPmdXfWjMa15XmuPfa3VUp20uI0Bkz1RycjKdOnXizjvvZMmSJRQWFjauR0ZFpQGpy6yifVs50iN1Z6pDFuyodVEczj9Mp9adWP231fQO6s258nP0DurNuyPe5dCZQ3Rf0J0xy8egOaehr19fpRSXzG85v/FB5gdcTBVCSXkJGacy6HtdX9p7tyeoVRCDQwcTHRHNk7FPYggwEDIsxGGJTnx8PH5+flWSNzVV1CU6DY+q0SpNFTlyYcaMGXWOTKvrrJi9bsvnTU9Ppyi3CGmr5FT7jx49yvLlyzEYDEx/aDqjw0Yzuvtopa3ZbLY1FFC8uZjR3W3vv/nymwx4egD4QNntZUqi1ZjoGLZlbmP+K/PpHdSbhP6OSVJTd6YqlcUQsGL3CmZNm9UomqjOKjYNVH1uvlyNiNorcU5X9rBcAfHVu16t8r7cr+jYaAdtC/AKYHT30dzX+T56Z/dm1ZOrHPI6vfnym8qxfDN9KV5XTFFukYPeZZ7K5OcTPyuRHu693NH00eAmuWHBolRGDH8oHIObAY++HgQmBxL8ejD0AS1a3DXu6G/Rc/MbN9MjvAcA7U3tKZlRQr+cfi7tf9m2jnomql7fERqinRq5fZWpyXty9OhRkZycLLp06SIkSRKenp5izJgxYuPGjcJisTSop6a+qLOLDUdzHqurqA1XbZenLXca4RG7IVas37tembGTX7eZ0UZ4TPYQ+sl6MWnTJGW7/e+QuSE2r3QCwj3JXUStiRKRKyPFdbOuE76v+wopURIkIjSJGhG5MlJErY0S3eZ3E8ZpRhG+KFy4T3YX2iStiFobJYS45EmOWhNV67E542pfVzUCpH7URttUjW5Z17wmmsNYL+fv3X5WbMuWLTUey9n71e1z8OBBMXnyZLFgwQJRWFhYbR+0Wm2VY2TkZCgRe/Z6XN057Z9N8v8bKwKkKc4qVqY53MO1pSZtu9b1uSVd65poTmOti73sTNvksWbkZCgRc640x8/PzxaZIUli4LMDlXPeOO5GW3SH/lKkh+dkT+GW6OYQNS3bzUQiJB/JFv0ReTEyxO7/bl5uSgQIFyOnXY07am1UraI0Fi9eLObNm3dFbNemQHO6h2tDddpWowPEnh9++EGMHz9e+Pv7C0mSRLt27cR//vMf8dtvv112Jy8H1bhuOJrzWGsbdia3nfv+XIe29mHK9r8zcjJE8KPBl8Q2EaGfrBc+U32U92M3xIqA6QGi7ZS2ggSE/2v+ioEs98tvqp8i6KFzQhVjOPK9SNFtQTcxb9s80XV+V+E+2eY4kft0OY4Pmcu9rlfSgXG5/WjO93Bl6qptqka3fJrDWOvzJdyZY2DLli3VHqs2xrf9sT/e9rFITk4WCxcuFEVFRQ7vyRqbMDNBeBg8RCvfViJhZkIV/c3IyRADnx0oAtsHOuhQYPtAAQhjW9tzY/3e9XUae2WHfH21v6lodXU0h3u4ttRF265FfW5J17ommtNY62IvO9Ng+2Uh8vKWsWPHOnVGy0tT3L3dHc4p+di2o7/ozOiJg51t/+OW7Oa4JObi/938Lm2XPCWh8dUIn74+Aj3CzdtNLF682EFL7ScVo9ZEOSxddzX2mTNnNopDuSlqdXO6h2tDvZbAOKNfv3689dZbnDhxgrVr19KvXz9mzZpFeHh4w4SjqKhcBnVd7uLv6e/Q1j5MOaF/Ao/3epzlI5YT1jaM4q+KbSW2toJeq0cjafDQeijVWR7v9Th9PfoSUR7Bja1u5B+3/IOf//yZh1c/TMrOFFuontZdOdf1xusBiFoXxYbDG9ifu5852+YQ5B1ER7+O9L2uL+PWjuPpz5/mi8NfcOjMoQb/vOpCU1nC0lT60VRRNVqlKVCf0F65asrZsLMOiT+rO5Z9dYHqzpW6M5Vlvyzj9Y2vExQUxGOPPYa3t7fDeeXlhbOmzaK0qBThJjgbdtbhvX79+tEjsAeb528m588cBx3qMaoHGl8NFXdWsOvkLpK2JgFVl6S4qhiTdyHPoUpZTZXLXKEmTm26qPqs0lSoi71cnQbbLwFMT0+vYp8lJycr5W8nJU1Szpl5KhN9J70tuan/xcYHsNnZnwI7Lp3DZDCxevRqBj4+EP8gf4Ijg/Ec6IkhwMBdUXfhOdATNz83nnrxKZ59/1n6x/YHdyg/X05ycrKDltonwvb38mfDoQ3Vamx8fDzu7u6NskxFtWevMvXxqBw+fFgkJiaKjh07CkmShK+vbz19Mw2DOrvYcFzLY61uxm3x4sUisH2gGPjswCrLXn45/otYvny5SEhIEA8ve1gEzAgQbWa0qeLBdp/sLjySPQSJiG7zu4nYDbHCOM0o/Kb6CU2iRmiTtEpCqfBF4UI/Wa/sWzmp3uWOta40FU+1GgFSO1SNbrm01LE609+axlrbKImPvv5I9EvsJ5IXJYuSkpJqj5EwM0EYAgwiYWZClfe4mHiai+HV9jokt523bZ5DBEjl2VNns6kNGQHSHGhJ93B9tO1a0ueWdK1r4lofqzP7bOzYsUKr1YqxY8falg+uiRL9nuonvPy9Li19kbgUCWL/fx9E0Jggl0vV7ZcUVhelV11Rg9pobGNd16ZiV9vT0u7hBlkCk5+fLxYtWiRuv/12odFohFarFffcc4947733xIULFxqko/VFNa4bDnWsl6iNOJaWlor7Z98vtIla8dclf1XWhs/bNk90m99N9FjQQ/i87iP6/reviFwZKcxLzaLr/K5i/d71DvlG5Fwhke/Z8oKs37teRL4XKUyzTCJ0bqjLcOraCqh6XZsnddE2VaOvDdSxVqU6Hfz555/FvxL/JW6bcZt49KNHXep5dXovG9Q6N50AxF/+8pdaj6Fy31z1Vb2uzZPaatu1qs8t6VrXxLUy1trkxZD11D/QX3EYm5eabZN6PnZODj0CnV1ll54XnSButjZao1ZEvhcpxieNVxzTzvrTmI6Ea+W6CtHyxlpvB0h5eblYv369GDlypNDr9UKSJNG9e3cxbdo0cfz48QbvaH1RjeuGQx3rJWpaI7njjx1i0MxBQpOoUcp22be3d3DEbogVkSsjhX6KXklwWhO1WaNZ27X26nVtntSkbapG22hJ17wm1LFWxZUObtu2TSQkJIghc4YopWxd6Wl1ehu7IVZofDXKORrD4Fava/OkOm1T9bllXeuauFbGWpu8GLKeBj8aLCRPSbh7uwtdb13VXB+VS93aOUckT8n2O1ISbkY3pWyts/7Uxg62py4afq1cVyFa3ljrlQPk6aefpl27djz44IOkp6fz5JNP8uOPP5KZmUlcXBzt27dvtGU5KipNgerWSF64cIEX3n+B74u+J7RVKBo0hPiFcOt1t5J3IY+0fWmMXj2a/9v+f4xdO5Z5P85j54mdlwKoL+P89mvI1TJa1y6qRquo2HCmg99++y0bNmygW7duxNwTQwdjB6W8eGUyT2WSV5zH0E5DHd6XtdY305cBjw0gsH0g8fHxxMXFkZ2dTVxcXJXjTNw4USmdXtN2lZaLqs8qLZHMU5l0ebALbm5uit46y2tk7mDG8pGFY+8dQ1wQlJ0vo2JXhS3HxwFsJW63YiuB63PxNSilbw33G9Br9VACmq80dBvZDS9/L55/8fkqfaqPHazm31Bx6QBZvHgxt912G6tXr+bEiRMsWLCAvn37Xsm+qag0Sd7f9T4hM0PYW7yX1t6tuSAuIBBk5Wfx7bFv+eTAJyRtTeJw/mHKreUUlRdhxYqH1oOnbn6Kvtf1pfv/dWfYymHVGsRhbcOYM2QOYW3DHLbbC7ea8O7aRdVolWuVyga3vQ4KIUhPT+err77ipptuYtSoUXzzxzfkXsilo1/HKnoKtoSjGw5vwN/Tn7C2YcrxZUfHkrlL+OrNrzh5/KRyDgAhhENfXCUurW9CU5Xmi6rPKi2R1J2p/Gr6ldYdWit2Z2VnQuapTJK2JpH/U77zg1RgS3RaCBwFngP6XHyvD/i86MOKKSvQu+kBsJRYyDydyQ2JNzA6arTiUE6clUhwcDBAFTvY3unszEGjTh6quHSA/Pnnn6SlpfHggw/i5uZ2JfukolIrKs+qpe1LI+K/EaTtS1Ne2zsaKrcvqSipdlbOmdF67tw5/vPpf8ix5vAnf3Ki+AQnik4gEFixUlRWRDtDO3w9fDG1MmG+3oz5BjOhxlDC2oQRHRHN0l1L2Zu7l88Pfl4vg1gVbhVQNVqlbjgzAuvTpingavZOCMFXX31Feno64eHhjBgxAq1WW2PFg8rvy8cvKSlBq9ViNpsd2s+YMQOTycSMGTMc+uLqPHWpuKDSMlD1WaUlImtZgFeAsq2yTSpXVJTcpEs79gT0F38kLkVC97/URCfp0Ek6hncZzvCuw5k+fTparRYEWNItHM4/TOrOVCZtnsT8n+Yz/fXpVZ4D8jNswmsTWLF7BdO/m85zLz1XpZ06eaji0gESEBDg6i0VlSZBZQdF0tYkh/KDSVuTHBwNldvnFuc6vM48lcmAZQPwnerLgKUDMHcwOxitBQUFLF26lEBLIBISPu4+3Hn9ndzf+X5CjCHoJB0GdwOB3oF888c3HCs8xsnzJwn2Deb24Nv55tg3RK2L4oneT2BqZcLb3ZtQv9A6j7upC3dz+RLV3FE1WqUu1CbktyHCgq/E339lgzslJQU/Pz98fHyYN28effv2ZdiwYWg0NhPHVTSdTFhbm3M6dWcqafvS6PJgFwLbB+Lh4YHFYmHT/zY5OMvtNdhsNqPVaundr7dSZrHyeWo6v0rLQ9XnlkdztW3kyb+0fWkOOmY/KVjTMj15gvHQmUPMGTIHgEETBhF0XRDHC48zKmUUtw2/DbAtf9Ht1IEOPA2eDHx2ID3+0QNexPajvXhQPUrkh36XHt+3fOmf15+4O2xLC2NiYli4cCGB7QPp90g/Hgp7iOiIaA7mHcQiLBjvMVaZDJSfYRkfZTD2prEgqFWpdJVrD5cOEBWVpk7lWbWE/gn0DupNQv8E5XW3gG7c1/k+oiOiq7QP8ApweJ26M5X039MpLCsk/Y900o+mK0brmTNnWLp0KcXFxRz1OIpAUFhWSO+g3nz8yMd8MuYTbgq8CYEgpygHrUaLwd1AVkEWq/esBglCjCEcKThCVn4WbbzbcK7sHEt3Lb06H14joq6tVFFpetQmcqwhosuuxN9/ZSdwcnIyBQUFFBUVsW3bNu677z4kSarhKI7IDvKkrUn8avqVMUvGKJEePUb1cLmEJT09HYvFQnp6urrMRUWlBdNcbRt7bbPXKPtJQVfL9GSnzzNTnnGYYMwtzmXLO1vI+TOHN6a+wVs/v8WoD0eReSqTNXvXkPtxLuKCoNRSyi2Rt5CRmwE7gNnAjdicHxaQpkt4/uqJ/ns9eTl5ZHyUQerOVMUp803AN9wz/x7u7XQvWyZs4fu075k2aBrhQeEsSlhUZTJQfoZNTpzMnCFziLsjjpjoGLZlbmuyk4YqVwfd1e6Aikp9kWfVZIZ3tYXNuXoNOLTX6/TMGXTpdXRENLtO7mLniZ1EBEUojpHc3FzeeecdLBYLjz32GN5/eJOQnkCvtr0wdzAzceNEoiOieaL3E0z+ejJ/7fpXLlRcwNzBzJo9a0CCuNttHu3UnamYO5g5nH+YC+UXFGdNSyI+Pp7k5GTV266i0oSIiYmp0QCsTZuauNJ//1arlQceeIBly5ah0+l47bXX6uz8ABS9N3cwk3403RbJMSSMmJgYMk9lKtEdlZHH+2Tsk5wNO4tvpi/BMcHEx8erBreKSguiudo2zrTNfru9rlXWONnp47/Zn953XppgDPAKYMBjA8j4KAOvTl4ceeMIB80HSe2USsGFAmV/6wUrU+dNtUV6bOVS3g932/9FuaDkfyXcEHED5/LO4dfFjxW7VwCwccZG9qXvg5tA87sG61mr8vmfnnKaU5pT0PVSXzNPZbIneA8bd2xUou3k7wlyhIuzCL3akpKSopxf1fYWwBWqRNOoqCUWG44rNVa5RGxGTkad2tZlv8qs37tedFvQTUS+FykycjKqjDUjJ0NErYkSUWujlOOfPHlSzJgxQ7zxxhti2bZlInxRuFi/d72yj33pxPBF4UJKlET4onCHfstlcOVj1qa8bV2pqaSXeg83Txpb264UqkY3HE1prLLujB07tsHLwgrheqzyeRctWiQ++OADkZCQILZu3VqlXW2eFzW1qesx6lOSUYimdV0bm5Y01pag0WoZ3IahJY7VlW1pP9bA9oECEO7e7sI/0F8EjQkSmmEagVSpxK1c9ranrcwtOttvw0iDcozA9oGKlmo0ttLjSLZ93fzclP4409iaSpnX1+6Wx1pfbW9OtLR7uF5lcFVUGpO6ZMWvLkyvpjWM9tvknCCfHvyUAe8M4GzJWYfzTP92Oit3r2RVxipGrx7NPW/fw32L72OvdS+HOh1i+vbp7Dq5ixe/epFx68Yxbu04hzwhlZfguAo7bIyEeM01NFNFRaX5IuvOihUrHPSnsdfKy+d96aWX2LNnD/feey933XVXlXb2zwtXfarpWVSbZ5V9GzVJtYqKSkuhupxzUVFR6HQ6OgR3wNjWSJmljLycPE5+dhLPWzzhARxL3PbBVvHlKFACkrdE6NRQVkxZweTEyQ5LV8LahvHII4+g1Wq564G7CL8/nNXfryYmJsalxlZnWzeE3a1qe8vC5RKYjz/+mP79++Pr66tsO3PmDEajUUksBnDy5ElWrlzJc88917g9VWlROAu9q0tb+7wdcrgcoPxfXupi/35C/wQmfjGRY2ePcbr4NH8W/amENps7mPn5xM9oNBo8dZ7sz93PXvaiQ0e5oZwTB0/Qo00Pci/k0sq9FSt3r0QrafH38lfOFdY2jE6tO5G6M5VOrTth7mBm6+9beaL3E2TlZyl9rrx0pyForqGZKvVH1WiVq018fDzjx4/HYrGg1WoV/alcqruheemll3jllVe47bbbuP/++52WF01JSWFV4ip6jepF9GPRDIkZ4rRPNT2Lanq/8nnkZTMq1zaqPqu0dFauXInVamX79u34vu4L32Fb5tIBzk85b2s0GOgDHloPArwCiAiKICI+giVzlyjPi6cHP018fDzHjh1zOP7y5ctZvnx5lfO6WqpZnW3dEHZ3QywRVWk6uIwAGTFiBPv371deWywW2rRpw65duxza/f777/znP/9ptA6qtEzqkhXfvm3l/aIjohnacSh5F/KUaAw5L0fmqUzF6yvn49Br9bhr3fHT++GmcWP6d9N5e9fbjPpwFHtz9xLoHUjftn2RkGgttWZkt5FMHjiZoZ2GcrLoJOXWcnIv5AJwnc91REdEO0SZTP9uOot2LGL6d9NJP5rOscJjivNDTuzkjJoycNdEU68Mo9LwqBqt0ljUNoJDztJvMplYuHChoj+NOVNWWlrKjz/+SEVFBb169aJv375O+5ucnEzOnznsX7ufsLZhLvtUmwox1b1f+Ty1pblWlFCpHao+q7R0bux/I0hwwx034K3zvvTGAaDk4s9W26bSn0r5M/lPIk5EkPh8omKvqtHLKlcLlw4QIUSttqmoXE3C2obh7+XPhkMblKot6UcvZeOXjdc1e9fw3u732H9mPzcYb+CBzg9wvvw8CJtnukJUAHD83HHKsstop23Hi/1fpJ1POzq17kTBhQIO5B3Az8OPG3xvAAE3+N6g1CRf9PMipn87nYsrHkHYEk4F+wRj7mBukDBrFRV7VI1WaSzqYpQ6c742lEM2NzfXwUlQUlLC8uXLWbduHYWFhbzzzjsu+1vZ4VHfPlV2VFR+XV9nj2r4t2xUfVZpKThz1pZUlNA5ujPd5nfjhidu4PSF00hbJVuSU7BVedED/cFNcoOtIM4KXn/tdYcl60/GPomfnx9FRUUkzkq8rIlAFZW6oOYAUWn2VF7b53Stn7j0++Z2NxN3Rxz+nv7E3RFHSmQKplYmJCQEgh/4gXO6c6zat4olvyxhwDsD2HlyJ1as6N30tPJoBRIcOnOIBT8t4NMDn1JqKeVY4TEKSgpo16odBSUFLPllCXtz97Jm7xpbpEqnoRw+c5hx68ZVEXh7Z8nloM4qqqioXC5NZa3ziRMnFCdBcXEx77zzDidOnOC5555z6J+z/to7PDJPZSp5m1wZ1/baaR+RV9lRUfl1bR0r8jETZyUSHByM2Wx2+hk703BV11VUVK4Wrya+SnZ2NpPiJym6mFucyxdZX3Ag7wC/n/0dCYnrhlyHxkeD9z3e8CJoXtTQ7d5udPujG5QBetCZdUx4bQL9wvqRkprC2bCzeHt7U1BQwBtT31AiqGvK6eeMukZSq7p6baM6QFSaNWn70ohaF4W5g7lK2auwtmE2w3ftOApKC7j/xvt59KZHGdl9JNO/m06FtYJDZw6RfjSdBf0W8BAPYZAMaLVaLpRfoNxSTkl5CaeLTyMJifCgcKYOnIrRw4ibxg13jTsVogJx8V9OUQ6bsjbxx9k/2JS1iYN5B5VokLC2Yfh7+rMpaxOr96yuEukhL5dJP5rusL2uAi0b5+PHj7+mRF19kKm0JK72/dyQS+ou5wt9u3btCGwfSKfhnXg99XVOnz7Nww8/TEJCgkP/YmJi2LhjI9/4f6M4mGXtH7duHNO/m87qPav5YM8HRK2LIm1fWpXk2bKRn5ycrETkTf9uOl0e7EJg+0AHZ0tg+0C6PNilWuO88mv5mLOmzSI7O5v09HSnn7GzyBA1WkRFReVq0WNUDzS+GrwGeimRygFeAYQaQ9FqtPRo04P7A++ndbfW6J7XcT7ivLJvp9ad+OOTP6AEjD5G/vmPf5LxUQZFuUVIWyWiI6IVTfUZ7IPVauXglwcVB4m9rewsUtpeZ6srkmCPvH3Sq5PIzs7m1cRXL/szutrPbJW6c8UdIBs3bqRLly506tSJadOmuWy3fft2tFotH3300RXsnUpzQRGwzZPYdXIXSVuTnLZL3ZnK6j2r+eLwF2QXZjOy+0iStibxQcYH5F3I48WvXmTZL8uYtmkag4MH8/Xfv8bHw4cyaxmZpzOVpTHh7cPZ+c+dDO86nLg74njq5qeYPWQ25uvN+Lj7YL7BzNRBUxndfTT3db6P0d1HM23QNJ66+Sni7ogDbJEpo7uPZnT30VUS6rnKUF1Xwzc+Ph6tVovFYmk0Y7kpCr36BaFhUPW5adCS7ufqvtC7ctTK+m4wGhi5aCTb227ns/zPyLoxizLfMqfnkbVedjDbv0bA6O6jCTWGcqTgiFKZa/q304laF8WyX5fRY1QPJSJD1uOCkgK+a/sd9yy4x8HZMmbJGH41/VqtcV75tXzM5198vtroGmfRLE0lIkdFRaXlkZKSQtB1QQyaMMhp9MSbL7/Js+8/y/xX5is59XKLc5k2aBr/uvlfPOD1AD2O9yCmWww3HL4BZgM7wIqVTw58Av1B46uhz8N9mDNkjlLxZfbrswlrG6ZoaknvEsLahnHk4yMODhIZZ3ayvc5Wft/V0nJ5u9cgLzS+GnqM6nHZn2FLemZfK7isAgOwePFiPv30U8C2dlGSJBYuXEi7du2UNsePH6/1ySwWC//+97/ZtGkTJpOJvn37MmzYMLp3716lXVxcHPfee29dxqLSxJE9tNER0Q7J4tL2pZG0NYmE/gkM7zq8VseSBezW625Fr9PzRO8nGLduHAiIuyNOOb65g5mP939MbnEuh/MPk7Q1iUNnDiFJEhISbbRtMJYZiWwfydixY3F3dyclMoUXv3qRIEMQrTxaYfQwKk4McMwmXbm/1b0OaxvGuw++63Q8rjJU17W6i2ykJycnYzabCQ4OJj4+ns6dO9dq/9rQ2BUe6sO1WgWnITVa1eemQ2PfzykpKcrxG/tv2Gw2s2rVKsxms7LNvnqMMx2R9T38xnBaHW5FL6kXQSFBfPL7J/ju9FW00v7ZER0RzeEzhzmUfwhzBzOdWncirzgPJIi7PU6JCEzdmUqoXyhLdy2loLSAIwVHCDGG8OYTbxL25qXn0pwhcxi3dtylvE52VFeZzNVvReOHQOLziS4/L2eVBtTqA82ThrahweaknjBhAhaLhejoaF588UWn7bZv306/fv344IMPGDVqVP0HodLisLfFAZ576TmKcos4/c5pUofatk//drqDdiqVFV9NZe6aucxcMJMPiz7k5KmTLC1Yiq+nL2cOnaFofZEtD8gmbAlQ+8P5vue5xf8WMj7KIKVHilM9k/vim+nLvAvzMBqNzHh9hsN3BWd2sr3OVn7fVQUv+bU50kz62PTLKo0rc63aoM0a4QJJkmr9o9FoXB3Gge+//17cc889yuvXX39dvP7661XazZkzRyxYsEA89thjYvXq1TUet0+fPrU6f33ZsmVLox6/KdGYY43dECsCZgSI2A2xDtvDF4ULKVES4YvCa32sjJwMEbshVmTkZCjH1k/RC/1kvcPx5e3uye4ifFG4WL93veg6v6twS3YT896fJ8YnjBcrVqwQZWVlIiMnQ0StjRJRa6LEvG3zRLcF3UTke5EiIydD2S9ype21fdv1e9c79KWpYDKZBCBMJlODXtfFixcLk8kkFi9e3GDHbEha0t9rddrW0BrdWPpc0zgagpZ0zWuiIcZqrw2NjatzVacjGTkZ4ql1T4kV768QU6dOFdnZ2VU0X4iqzw77Z4x8/LFjxzqcZ/HixcIQYBDeI71F1JqoarXb2TkbC/Uebp640rbGsKErKipEaGioOHz4sCgtLRU9e/YUmZmZTtsNGDBADB069Krb0C3pWtdEfcZ6Newpe52M3RArvEd6Cy9/LzHw2YGK5ukn64XHZA8Rvihc0b+MnAyBZHMJz5w5U3gmeQoSECTa/fS0LwWAwAfh/Zq3MAQYanzmZORk1KrdlUa9h5sv1WmbywgQq9Xa0lkSRwAAioZJREFU4M6W48ePExwcrLw2mUz8+OOPVdqsW7eO//3vf2zfvr3B+6By9XDljU3on6DM4tUWZ57evAt5IKrOysnb5ciQNXvXkJWfhdaqpX+3/owcORKdTqeETCNg4+GNnC4+zb4v9rFlwhb87vXjWOdjHDxzkI5+HQGUthmnMzhWaKtf7iyK40rOttrTWB5p+ygT+9cqV5aG1mhVn68druRslf257CM2qotqaCu1pdPhTuhMOh5//HGCgoK4jusUfZVnMJ/o/QRLdy0loX8CmacyySvOY2inoURHRDPwyYHk/JnDipUrwIoSaZKcnExRbhGGrQZGThmp5F1yFaEoI79v7mAm/Wi6Q7ua9lW59mgMG/qnn36iU6dOhIaGAvDwww+TlpZWJUpv/vz5jBw5UtXoZsDViKitYotHQ/Rbl7QrOiKavOI8fv7zZ44UHCF1ZypzhswhdWcq0k0SYrcATyg5VwJSpYMfdXzZc3RP7o64G98XfVkyd4nLZ05KSgrPvfQcZcFlGDCokRQqjU61S2AaGuGkBJgkOf71xMbGMn36dLRabbXHWrx4MYsXLwZQEoo1FkVFRY16/KZEY491uH44p/ecJn3PpXP44svsLrPhJKSfrP+5/+73d4Aqx6+8fbhmOLd1uo1Az0D8jH58++23ANyru5deYb0A8HLz4nTxaUqNpYj+Ap2bDm2Q7Z5sa2lLYWkhb3R+A71Oj4+HD3kX8mhPe6efXUlJCbGxsZSUlFzR+6hz584sX74caPjrerXGVBuupb/XhqQh9RlUjW4sGmKs9trQ2J+b/bn2ZuzlUZ9H+TPjT5daX1FRQV5eHiaTCS8vL/bt28e+ffsc2mQXZtPhQgf8yvyUZ8cvB36hT3kf/D39Ob3nNMnTkjmXdw7cQVOuIfi6YNLT05kzZw4nTpygXbt2nDtwjutzr+fL/V/i3dqbDu4d+GXbL5z2Oe1wnl+2/QJAhwsd+DPjTzpYXLeTt9UV9R5WqQnVSd3yuBpLJypPIFaeuJOXa1deKhMdEU1ufC6///E7fn5+BBcFc6L0BJ46T4rKimzlAPoL+OTigfTQ675eTpf+ZZ7KZNLmSRzKO8TUQVMvOaYxsC1zGwATN06s1iE9/bvpVZa9X60JR2c0pb6oVKVeDpDPP/+cPXv20K5dO0aMGIGXl1et9jOZTBw7dkx5nZ2dTfv27R3a/Pzzzzz88MMA5Obm8vnnn6PT6fjrX//q0O4f//gH//jHPwC4+eabHdYXNzTp6emNevymREseqxCCzZs3s/uX3fTu3RsPgwdpJWlOBVYWft+KS17rb9y+YfWe1QR4BpB9Lhs3jRudWneipLyE40XHeaj7Q4zsMLJKPpODBw/y2rTX6DGqB3c/dDfAFZ8tbOjrevDgQWbMmEF8fHyTu19a8j1cW+qj0Q2pz6BqdGPRnMd6dt9ZRR/NXc1V3j9+/DgrVqzAzc2Nxx57jN27dzsdq6zPgyIGVYnCkLe1OdWmyjp2Z8fpF9aPotwifHx9sLpZef7F5xn7/NgqxwQcIkCqO3d9DN/K17UlG8/N+R5uCOprQzfHScRrydlVn7FeSWd0fbCftLxQfoG7LHdhbWXFoDXwfJvnKdeX4651x2K1IBBYb7TCMKAI9EY92iItb857E08/T65vdz0AucW5VFgr6H++P/0r+nMk/QjPP/88YLNDTu85XaNTObswm96lvUHg0KYxJufqew835YlCV1xLf68uc4DMmzdP3H333Q7bLBaLGDJkiNBoNMraxRtvvFGcOnWqVmtxysvLRUhIiMjKylLWL2ZkuF5bq+YAufI0h7Fm5GSIqDVRImptlMPabFdrKeU1jQs/WigSEhLEp59+KqxWq1iettxpThIhLq2RtF8jHrUmSugn64XPaz6CRIRHsodwn+wuNIkaoU3Siqi1US7zmVRec+nqvI1Fc7iuDUVLGmt12tbQGt1Y+lzTOBqClnTNa+JqjLUuuTDqmzfj999/F6+//rqYO3eumDNnjjCZTLW+v+p6fvs28nPDaDQ22Nrz+uRYqXxdr2SelitNS/p7daVtjWFD1yZPU4cOHcQNN9wgbrjhBuHt7S3atGkj1q1bV68xNAQt6VrXREsea2lpqbhn9j3CK9FL6P30YubMmQKfS7k/3JPdhd80v0u5QCIRGl+NQG/LBaLx1TjYv1Fro4S+td72nkZTRetq0nH7PHy1+R5wOdT3ujb1fHnOaGn3cHXa5rIM7rp16+jSpYvDtrfffpsvvviCRx99lF9//ZX333+fU6dOMWXKlFo5W3Q6HQsWLODee++lW7du/O1vfyMsLIxFixaxaNGi+vpwVK4xpn87nZW7V/Jh5ocO5a1eTXyV7Oxsnn3xWcatG0favjQmbpzI9G+n8/bPb7Ns9zJuvfVW7rvvPvac3kNJRQltPNvgqfOk+4LuDFs1TCkBFh0RzdBOQ/nuj+/4v+3/x6gPRzGy+0gGhw7GR+9DqG8ob9zzBg+FPcT9ne+n74m+fPn0l/TL6UfXgK6YWpkcyonZl+dyVfK2JdIUS+a2FBpao1V9VqmM/Pc74bUJTssJOsNV6cHqOHLkCMuXL8dgMPDEE08wa9YssrOzyc7OdqkfmacyGbd2HOPWjSPzVCZp+9KI+G8Ej7z0CPMenseE1yYo5XQrl3a07+Ntw29jVMooJrwywWWp2crnqknXGqJsrVr6tnnTGDZ03759OXjwIEeOHKGsrIz333+fYcOGObQ5cuQIR48e5ejRo4waNYq33nrLaYSeikptKS0t5b333qNTYSdGdBxB74d6gxZ0ZtsiAn+9Pw+FPcQd19+Bm8YNN40bmq81WM9a0Wq0+Af5M+CxAQ72b9ztccybNg+TycQjjzxSRevkZTquoqTD2obx7oh3effBdx3axMTEcOzYsSYRNdeU+qLiBFeekXbt2on33nvPYds999wj/P39RWlpqbLttddeE506dWoAP039UWcXG47mMNZ+T/UT+CCCxgQ5eH4HPjtQ4IOQIiWhn6IX4YvCRcD0ANF/bn/xl4S/iCUfLxFWq1UIYYvImLVyltBP0Ys2M9oIEhHaJG2VCjLuk92FlCgJ98nuInZDrMsID/vZuqsR4VETW7ZsuSre6Ksxi9kc7uHaUp22qRp9iZZ0zWviSo5V/vsNbB9YrwiQ2sziPb7qcfF08tPi//7v/8S5c+eEEJdmzmbPma2cvzJRa6KENlFbRZt1Rp1Dn+XZRrkflSt6VdZrZ32WqyLop9iqjDWGrqn3cPPElbY1lj5/9tlnonPnziI0NFRMmTJFCCHEwoULxcKFC6u0bQpR1C3pWtdESxxrcXGxSElJEUlJSUpEaEZOhlietlzM2zZPqa4ob5e1M2FmgjAEGETCzIQq78ms37veYf/Gpr42cEu8rq5oaWOtVwTImTNnMJlMymuLxcK3337LgAEDcHd3V7bfcsstZGdnN5Z/RqWF42yGztWsncyRj49AIYitwsHz++bLbxK5KJKQu0MYHDqYwaGDKSstwzPfk/F/Gc9ut93sOb0HsEVktPZszejuo3n1rlcJMYZwg+8NmDuYleOZO5gJNYbS/4b+PNT9IaIjoknon0DvoN5VKtbYz9Y11QgP+2zjVwp1FrPxUDW65dHUIqbkv9/JiZOrnY2zx37mrqZokDc2v8GH+z9kj34Pjz/+OAaDAbg0c+bt743GV0OPUT2q7iyBTquj9Z7WrHpyFf1y+tE7qDcxE2OUPstajEDpR9LWJPbn7Se7MJuwtmFV9NpZn6MjohndfTSju48mOiJa1TWVGmksfb7vvvs4cOAAhw8f5uWXXwbgqaee4qmnnqrSdtmyZYwaNeoyRqFyLVNcXMw777zDiRMnlIhQsGm8ycdEVn4WxwqPKZW07JOrng07y7bMbSQ+n0jmqUyi1kWx7NdlDrqatDWJXSd3EfNJDJmnMmu0/S+XuLg4srOziYuLa5TjqzQvXCZBDQwM5OTJk8rrnTt3cuHCBW699VaHdlqt1kHMVVTqgmxswqVM1M622TM5cbLTrNlhbcPo6NeRH7J/IMQ3hA9//pBCSyFZrbLYJe1yOGZY2zBOG0/zrvldALLys1ixewXpR9OV5KXpR9M5ef4knm6eLLhjgbJt+YjlVb4I3Db8NgYEDOAb8Q23cZvTfl9tGiLbeF0T81VX6lLl8lA1uuVxNUoiVsfl/v26Kn0OkJGRgeGAgdu8b2PaQ9McEkEqJWfbmHn2/Wed7h93u82IXTN3DcV5xXyS8omSxPet+LeUdnOGzCHzVCb+Xv6YO5g5nH+YC+UXFCe2vdFeuZSujFwVQXkdY9N/tRS4iitUfVZpzhQVFfHuu+9y5swZxowZQ6dOnaq0caXvlW341J2pHCk4QogxxKFtQv8EYj6JoaSiRHGMVGf7Xy7iYhJh4SSZsMq1h8sIkL59+7Jw4UIqKioA29pFSZK47777HNplZGRUqRSgolJbnEVL1BRB4WpdnWy8/qX9X9ixZwddS7vSxacLT9/xtINRK3uZSypKauxHiDFEqYNe3Wxm6s5UVu9Zzeo9q+u09v1K0hDrEa9GFImKc1SNbnm0tMgCZ+u4U1JSCAoK4qWXXqLP9X349JlP6RPcx2E/WWvPlZ1zGXkS1jYMf09/JLOEIcBQ7Wcm9yP9aDo/ZP/AkE5DFEd35fNuOLwBf0//GqNdVC1UqQ5Vn1WaK4WFhSxdupSCggIeffRRp84PcJ2nw96ezjyVSd6FPO4MvpMebXtw6MwhJcpjeNfhbHlsC0+GP1mr/HiXGyEyY8YMTCYTM2bMqFV7OSIzNze3XudTadq4dIC88sorbNu2jeuvv56wsDD++9//8te//pWuXbs6tFu+fDn9+vVr9I6qNG+qE668C3lM/3a68p79jJz9PpmnMhm3bhzj1o6rEi4nh9it37+erYe3suP8Dvw7+rPvuX1k5Wexfv96Nh7ayKEzh0jdmcqyX5dxpOCIwzmjI6JJ3ZnqsG35iOU83utxoiOiMXcw08azDYfzD1c5f+UQ6ZZKS/uC1pxRNbrlIBtaQLNPmlbTMp5XXnmFnJwcvv32Wx599FE8PDyqtJEN4QCvgGqPFx0RTUx0DNsyt9XqM6vJwJbf9830rXEpkqqFKtWh6rNKcyQ/P5+lS5dyuPAwWZ2zKPYurtIm81QmRwuOMmzVMMUet8feMTLhtQm898R77Pp8FxsObSBpa5LDRKJ925oSn9YnwbY9dZ0ElJ3cJ06cqNf5VJo2LpfA9O7dm6+//pr58+eTn5/Pww8/zH/+8x+HNsePHycwMJBHHnmk0Tuq0rxxtaxFjpxAgL+Xv0MoctS6KI4UHAFshumoD0dxIO8AOo0Ofy9/4FK4HNiWsYhyQZkoI9Q3lEn3TgJsuTwW/byIc2XnSNqaxPIRy9n6+1ZKK0pJ3Zla7dIbe2eMHMZ35OwROvp1dDj/nCFzeHfEpRDploq6pKXpoGp0y6GpLX2pjpqWwVU3lm3btnHLLbfw/fff89prr7kM/VciNtLTqz2evT7Xhpray+8HxwTXeD1ULVSpDlWfVa42slY/GfskZ8POEh0RXW1kW15eHu+88w7Zpdls9t3M74d+57uc7+js3xkAo4eRkd1HkrQ1ibE+Y/n84Oe4adwAm/3u7PgZH2VgPWuleHMxUWOjMHcwk3403cEJXdul1dUtqWwM5GXj7dq1uyLnU7myuHSAgC2E7913XX+pu+666/jkk08avFMqLQ9XwhUdEU3ehTwQju9VXjOYujOVrIIsJEmio19Hh7bmDmY+2P0BXaWu+At/jhiPMPXeqYAtgiSvOA83rRutPFop6757tO2BUTIyOGJwjX106GtxHkhUWSrTkqhrng+Vq4eq0S2DhsjPc6WoyVnjaizffPMNM2bM4Pvvv+f111/nn//8Z63OV5fPJm1fGi9+9SKd/TszdeDUWiVtrUzmqUy6PNiF8o/Km8X1UGm6qPrcsnFmK8n5i2pyNjgj81Qm07+dDpItx1FN+9d0LlmrZ02bhf4FPeA6t8apU6d49913EUJwvst5sg9mo9fpySrI4uCZgwC4adzIOJ3BkYIjeLT24L7O92H0MILkOneHfc6+mCG2z6jy8sPaTgDU1eF9uchO7vT09Ct2TpUrh8slMCoqDYmr0DZXtbyjI6J5vNfjSsLR6IhoHur+EI/c9Air/7baIVxu08FNfLz3YzQlGtp2acvpstOkH02/FNEhwZPhT7LlsS0M7zrcts770AZ0Gp3DOV31UQ7B/j7te9598F3eHfFurcL1mivq2nYVlStLQ+TnuVLUtPSj8liEEGzZsoXNmzfzww8/cObMGaZMmaK0r7y00dnxNu7YyJ7gPQ7LIZ0tqUzamsS+vH18fvDzeodJp+5M5VfTr4xZMqZZXA8VFZUrQ+XleM5spctZplGXXHKuKqvYI2t11DNRDsu35f1lDT158iTLli1DkiQef/xxnr3jWR7v9TgpkSk81P0h7ut8H/d1vo/R3UfzRO8nCDGGYGplYurAqfh7+TOy20iXSwvtnweudLumZ0pjV4dpDJpaRTeVqriMAPn73/9e64NIksSSJUsapEMqKlDV01s5C79MYWEhvod96UUvXo18lWBTMMadxioRGpWdKwABuoBa9aU5hac3BPHx8bzwwgucP3+elJSUa2LMzRFVo1WuBnVZ+iGEYNOmTXz//fdERETw2muvMWXKFAdDd/p301m5eyVajdZhGaQ99ssToyOiHZZH2rdP6J+gRIDUNzKvukjAy5ndbYzjqDRdVH1ueVS2BZ1Fp13OMg1XUcbOcFVZxR5ZqydunMiRHUc4UmBbvi1XZlmxewVF54voeLAjHh4ePPbYY7Ru3Zo2tFF0tXK0xsSNEzlScIQ//f9kzXdrSNufxnu73yMlMsWpltlrnaul8DU9Uyrv1xyilK+17w3NEZcOkC+//BJJkqrd2Wq1cuLECVW8VS4LWSBD/UJZumspCf0TnGbor0xBQQHvvPMOfxT9wVm/s5R6lVYtaXghj0lfTcLoaVRCCu3Xl9eG5hSe3hDExMSo4t0MUDVaBZrukjUhBJ9//jnbt2/nlltuYejQoUiSpCx9kfttesCE9jotocbQapcfyr+rM/yHdx1eq2dHdVQXZl1Tifba0lDHUWm6qPrc8qhsCzr74n45yzRcTfQ5w14Ta3KiOltqHh0Rzblz5/De542nwZPHHnsMo9FYq/PKOfTQgIfWg9PFp0namuSyspa989r+d22pvF9zsE+vte8NzRJRT9LS0kTPnj2FJEliwIAB9T1Mg9CnT59GPf6WLVsa9fhNiasx1tgNsSJgRoBoM6ONkBIlEb4ovNr2GTkZ4qFVD4k+U/qI2NdiRY95PZzuF7shVuin6IU2USvcJ7uLbgu6iag1USIjJ0MIoV7X6li8eLEwmUxi8eLFjdOhRqQlXdfL0TZVo1smlcdqMpkEIEwm09XpkBMsFotYv369SEhIEF988YWwWq1V2sj9DmwfKGI3xCq6bI+z65qRk+GyvRCNq101nftyjnMt38PNmfpq27Wizy3pWtdEcxrr4cOHxZQpU8T8+fPF2bNn67RvRk6GWJ62XGTkZIj1e9eL8EXhYv3e9S7bNoRm2nOl7dPmdF0vl5Y21uq0rc45QP73v/9x6623MmLECPR6PV9++SX/+9//Gt4zo3LNIJcefPWuV+kd1FtJVAqOa//kNXWPvfQYH+7/kB0VOzjW4RhTBk+ha0BXTK1MDmsE5dK093W+j45+HcnKz6rV2kqV5pWTQMURVaOvLRq6HGt91i7b67TVamXdunX88ssv9O/fn8GDByNJUpV13HK/n3ruKZfHLakoqbL2u6bcS42Zw6ih8j611PxRKjWj6rPK1ebgwYOsXLmS1q1b8/jjj+Pj41On/cPahmHyMRHWNozhXYez8587XUbeNYbWqfapSkNQawfITz/9xKBBgxg0aBBFRUWsWbOGH3/8kUGDBjVm/1SuAWSBfOYvz1QRUvuEUrJh+9vq3xAINGjw8vRieNfh3NvxXn44/oODc0NOsPrxIx+zevRq/hb2N0Z3H93iqraoqICq0dcqDW0M1seBIOt0yo4UVq9eze7duxk4cCADBgxQlgFUTg4o9/ts2FmXSQNzi3PrnFCwoR1CKioNgarPKk2BvXv38v7779O2bVsee+wxDAbD1e6SispVoUYHyO7duxk+fDi33norR48e5d133+W3337jr3/96xXonsq1jhwdEh0RzbPPPouvry8DBg/gb13+xqM9HyXu9rgq7ZzhqtqMPZmnMhm3dhzj1jmvRmDfrq4ZqRszi7WabfraRtVolYakPg6E6IhoHgl7hPYn27N3716GDBnCnXfeqbwv52Ma2nGo01LorrQ7wCuAoR2Hknchr9baKTtWbht+W701tz56reqwijNUfVZpKuzevZvVq1fTvn17xo0bh5eXV632q60e1taGdoasn4mzEptdxReV5olLB8ihQ4d45JFHCA8PZ8eOHfzf//0f+/btY+zYsTUmdlJRaSjk6JBWJa2oqKggKSmJ91Pe54OHP1DK0dq3q8654UpU5femfzedD/Z8wMrdK5m0eZLL9vUpc3Y5pdFcIT8w4uLi1LK11yCqRqs0BvWJKOls7Eyf3D4U/17MAw88QL9+/ZT35HKNnxz4BH8vf6canVecx/TvppO2L81Bd/U6Pf5e/mw4tKHO2nm55Shr2rc2JTFVrl1UfVZpSvzyyy+sXbuW66+/nqioKPR6fa33ra2W1qWMb2Vk/Zw1bVaD28oqKs5wWQWmW7duWK1W7rnnHp5++mk8PT35+uuvXR7o7rvvbpQOqlw7ZJ7KZPq300FCqdgCcOTIEVatWsUR9yN8LX3N5o2beTL8SdKPprvMgJ22L42krUlKRRlnWfdPnT9F2zfaEtYmjIzTGQztOJRQYyhZBVlknMpgU9Ym8i7k8e4Ix6zc9clkfTml0VwhPzCMRqMyY3u5FSGaakUJlaqoGq3SFHjrrbd45ZVXuOOOO0hOTmb79u2MGDFC0RBXVVvk6l95xXms3rMaJMg4lcGxwmPAJZ02dzCz9fetmDuYHfarrP2ZpzKZ/t10EDCy+0iXESfVkXkqkwmvTWDXB7vo/VBvoh+Ldnm+2pTEVLl2UfVZpamwfft2PvvsMzp27MjDDz+Mm5tbnfaXNdTcwczEjROrrdYll/Gt3Lamst+yfkbGRLLNZ5ui900J1T5uWbh0gFgsFgC++OILvvzyS8BW1s4eSZIQQiBJktJeRaW+yN5jJPD39GfOkDkcOnSI999/n9atW7Od7Rw4fYDDBYfJLsyuYijbk7Q1iV0n/3979x3W5Ln/D/wdQEVENriwUkBBo6iodbZgrVasinsypIXW2qXtqdpWZfR4qm091nP6ba3gD8FVi4u2Cq4aWvegDpaCgoIiI7K34f79wUlMQhISSEISPq/r8lKSZ9wPkXcePrnHDdHSXLIKEI8rHqOougjJ+cl4c/ibosAd1WcUSmtLkV+RD7BmhwbXgStajlE80OXdKAv3Ufdyh+I33MIwtrGxQUlJCdasWdOqgNaH5cVIE8po0t5qa2uxbt06lJSU4MqVKxg2bBhmzJghkSHylmsUFqV9XH0wf9B8gAOM6j0KP1z9AfdK7ol6gfByeMgtzwUvhycqZu+6uQtJD5Kwe/Zu0bH4Nf8rpDAgpaipkOI3xE+lyfeikqNwNuYsGssacefwHXC3cbEqcZVE8Vx4E+zt7Q0ej6dwSUzScVE+E01SdL8p7uLFizhx4gTc3NwwyHsQ3vr9LYABayY0fcgo/Uu9dCH5UPohlNaUwqqrFQ6lH0JCVgIAwNe0+aSnwmV8IyMj4feyH5gXA/53y93Sst/C/FyVuAq5t5/nvS6h+2PDIrcAcvbsWW22gxCJ6nGwZzAyMjIQFxcHe3t7BAQEoM/DPlh7ei362/bHZOfJiL4RLVElFn9DCPUKxdrTa+Fo0bQyjHQBIrUwFcYcYwCAZy9PbJ26FasSVyHhXgL8hvhhzfg1cLF2kVvpltWjRNZjbaWo4izrhlt4gyV9o6Us+hRTf1BGk/ZUXV2N3bt34+WXX8aVK1dEQz+kM0Re8VfYs2PuwLnwdfcVDZW5V3IPmU8zYWVqhTet32xWvA72DEbSgyRkl2aLepDsj9kPk3MmGLZgGPpP7o+5g+aKegiqItgzGLcDbyPlYIqo/dLnF94E83g85ObmtuI7RzoCymeiSfLuN8XvGd3c3PDHH3+Ay+XC7WU3LDq4CHef3oWxkTFszZo+ZJT+pT4qOQq/pP6CZ4JnOJd7Dg/LHoIxhs4mnTF/4HzRfE1FaUVy2xYREYHK4kqYJ5kj+IfnGaxMHmuit7S60P2xgdH8Kryap8k1zBkzvHWRFdHUtaq6bvft27dZeHg4i4yMZNXV1c2eX5mwktl9bcdWJqyU+5isbcS33bp/Kxu+fbhofXJV1iuXta0m1jt3dHRkAJijo6NS28v7PtP/Yf2k6WzTFspo9dH2tcrKtSvZV9jETRPZ+xHvsx//+JEN3z6cHU0/qtJxZeW11SYrZr3JmnX+sjPzP+Qv91rF2+R/2J/BAirlZFuo+l6mLPo/rJ8MIaM1eQ3t/Vpr6udVFuG1auJeUOho+lH24ncvsj7f9mEz9s1odg7hPaO9vT0LDQ1lhw8fZgKBgK1MWMk6R3RmRmFGzP2/7qL9pL8/KQUpbOD3A1nnLzsz563OrPOXnVmfb/uwgd8PZP+59B/RdYlf66QPJ7EevXuIjqHN77k2tPf/YW0ytGtVlG1KL4NLSFuoMkHcjRs3cOjQIRj1MMJZq7OYdmAaXL5zwcz9MxGfEY+Z+2fi14xfMdh+MPjVz1cHcLZ2BgccdDXpilWJq+Dt5A0fVx+JbYD/rUhQzYeVqRV2z97dYhdpWROoypp0VRPrnau6IgOtj04IUSfpCfDKy8ux+ufVuFx7GdVu1dhxd4douKEi0pOGSq/+EuwZjGVDlyHaNxorRq6Ac7Yzbt++LXNVFfGsXTN+DSYtm4QevXtI5GRbVmVRtC9lLCH6oz0mJ9bEpPdC4UnhyC7NxqPKR8grz2t2v7l+/XrY2dlh9OjRGDFiBGbNmgUjIyMEewZjIXchlnosxcEFB0X7jfMdh3mR8zDOdxyApmyNmx+HFSNX4N9T/40VI1fgVedXkV+Zjy///BI7/94J/yP+qH1WK7rWszFnUfC4QPQ9Fs9IWgGR6Cq5Q2BUDQvqEkQUUbbr2LVr1/D777/D2dkZ1+yu4XDyYdQKmoL2QfkD5JXn4VbBLQiYACW1Jbj06BLAAWJnxyL6RjSKq4ux/fp21DTUNA2nAZrGhaNpzKOwy3TCvQRMcZsic0w6INmlUBNDW5RF48qJPJTRRBvECxQlJSWIjY3FUMFQuAxywSqvVch6miWacFoR6a7W0kNjxL/2dfdF35C+WLlyJb7++muFGch14OL0ttPANsXna4l413Ea603aivJZN7THsAVNDuMI9QrFqhOrAIZmmZtSkILjguOY/958zBg9A1OnThWtOCScn0OarPtb6SxOLUxFSmEKsp5mwdTEFNml2SiuLhZdo/iwQeFQdG8nb/ByeODX8JGQlYDbCbdx5/AdbNiwAeN8xyk1f0lLKKdJW8gtgISFhbW4s/hSXhTeRBF5v8jHZ8Tjs9OfwdXWFYvsF+Huubso71WOf5X/C70ae2Gy82RU1Fcg82kmOht3RtCwIJy6fwpZ/Cz0MO+BS3mXRBOVCuf9qKyvRFVDFSD878lp+iM+6Z7fED/YmdhJtEXem5Yuj0kkHRdlNFGHlma2F94M8/l8REdHo6GhAZ8u+xR9+vQRPd/SZHWRkZGoqqqClZWV0v8PN2zYgNra2lb/v1X1Fx/xm2ka603aivJZN7T1Q6TWrPyhzKT30oUC4d8tFQVcbVzh6+bbbLvGxkZ8fvBznCk6g6m9p0oUPxSRdX8rPcEq14GLUK9QhCeFI2hYEO6X3Ed3dBet8iJegA44HIC4tDicyDqBopoi+Lj4YGjeUJz9/iwaGxsRERGBeX3nqeVDRcpp0hZyh8A0NDQo/HP16lVMmTIFjDG4urpqs81Eh8jr3ibrcfHHhP/+7MxnSOen49jdY/jPuf9g0KBB4DEe7vDv4K+Hf8HFxgVnl53F/EHzUVFfgfsl9/Hr4l+R9n4avp/2PZaPXI41E9YAaKpUT3WdisqGSnDtm7pGrxm/BstHLMea8WtEXa7XjF+DrVO3wtREch10eUNYNDG0hZC2oowm8qjSNViZLuKFhYWIjo6GQCBAYGCgqPihrIiICJSUlMDc3FzpXyJCQkIwZMiQVv/youpQFfHhhjTMhbQV5bNh0NQQGuEHcuFJ4RJ/tzRsRtbwGoFAgMOHD8OxyBHT+k5DuG+4zOJHS8NRhM9vPre52TkOpR9CenE6rj6+iq1Tt6KivkJ2e//3gaOrrWvT/faENbhz+A4aGxvB4XBQWVkJy1RLieGPrUU5TdpCbg8QY2NjmY9nZmZiw4YNiIuLQ58+fbBjxw4EBQVprIFEt6kybET4GL+aj5SiFGSXZuPlvi+jpKIEtbW14Pbi4kL3CwjqF4TSi6UAp2leD+F8HoBklVpWlV3WkovCJWu9nbzBr+Fj87nNoqIJIfqKMprIo0rXYEWfokVGRiI0NBRjx47FuHHjEBgYCHt7e5Xbo4uf1El/skvDDYk6UT4bBk1ll/BeVVYPEGX2E/797NkzHDp0COnp6Vg6eSnGjx8vd9+WVi8EINFLWqItTOwPADszO5lFjDXj18C2q63EPbjwe1hZWYnS0lLs/G4nraBF2p+yM6k+fPiQvfXWW6xTp07MwcGBbd26ldXV1bV9ilY1oBUG1EfVa5U327WiVVL8D/szq01WbPj24WzH0R1sdOhoZvGlBRv+43DRqgDCFQKGbx8udyUXZYkfy/Sfpsz0S1O2MmGlzr2umpw5W9euVZMM6VpVyTbK6I5BmWtVV5b06tWLAWCWlpaMz+c3e16Y6UfTj6q86oEyKyUoswpMa6m6wpam0f9h/aRstnXUfDak17ol2r7W+vp6tmfPHhYaGsouXbokelyV+/Kj6UeZ+3/d2Yx9MxTmuPS+rblWee9LqrxfaXKFHXno/7D+UpRtcnuACBUWFmLjxo3YsWMHTE1NsWHDBqxatQrdunXTfHWG6Dx5Yx1lPS58LD4jHucfnMfjp49xJv8MXnN9DfVV9aKxheIVZWWr4oqIV9oPpR8CGFpcx7w90IROpDUoo4k0dfRmePjwIXr06IEnT57g9ddfh42NTbNthJ8eJj1IQm550yd6yo7pbsvk0uqYmFoXe6UQw0P5TDShvr4e+/fvR05ODmbMmIERI0aInpOXj7Luy3k5POSU5iCnLAcu1i5y81SZeU1aIu99SZV73/ZclIAYFrlzgJSVleHzzz+Hi4sLdu7ciY8++gj379/HunXrKLhJm5zNPovssmwU1BXgWrdrqLKrQm55Lu6X3BfNtSEMW1933zbPvyF+rNjZsYidE6uT83mouuStNtAyY7qLMppoyv3797F7927k5OSAMYYLFy402ya1MBX8Gj58XHwQ6hXarDt0S+PNpZfBlUfWcZTdV5GWxo9rcvlGYvgon4mm1NXVYc+ePcjJycHs2bMlih+A/HyUl6VjHMegq3FXOFs7t3hu4TGEy+Cqgyr3vuLXRvenpC3k9gB58cUXUVZWhilTpmDdunXo1asXSkpKUFJSInN7Z+eWf3CI4ZOePVr66yNpRxB1NQqMMXQ36Y5vp3+L/rb9AdAqK7o4Bp16peguymjSFsI5MLy9vcHj8URzYWRmZuLAgQOwtbVFREQEvv76a5k3plHJUUjISoDfED/4uvs2WwmmpU/qlP1EsaVlGjWFPmkkbUH5TNRF/D7aubsz9uzZg/z8fMybNw9cbvP7bHn5GJUchV03dyHpQRJ2z94t2rasrgyldaWIvhGNV198Ve4StZGRkfj484/BvBhGBIxodvzWUuXeV/zapkZMpftT0mpyCyClpaUAgBMnTuDkyZMtHkggEKitUUR/Sd80in/97eRv8fGvH6OqsWmJ2gY0YNbAWaJtie6hbuK6izKatIWwuLl//34IBAJERERgwoQJOHjwIBwcHODv7w8zMzN88MEHMvdvaXlwdS0f3l7LkNPy56QtKJ/1T2uWvFU36eVxgz2Dsfn8ZsSlxaGgogCv8F9BUVERFi5cCDc3NwDA5nNNz/Or+YidEyv32MGewUh6kITs0mxEJUeJ7ruFS9yGeoUqLPxGRESgsrgS5knmsFtup6HvgPLo/pS0hdwCSHR0tDbbQQxEsGcw+DV88Kv5SC1MFd08djbqDMuNljBtNIVFJwvUNNZg1ZhVEvumFqZi8/nNAINolRbpSrT0NuIV6tTCVGw+txngNM1Era5hLrrwpiiPptumi71SSBPKaKIMeRkhvHkU9gB58803ERcXh969e8PPzw+mpqYKjtoyrgO36eZdQSaLf3IJQOa24p/4SX/S2dI1tkTe8aTPq4guvz+Q9kP5rH90ocerrHmVwADGGO5l3QO3kYslS5bAxcXl+U4csT8KcB242D17t0TmApDowedq07Qks6zCr3jBwdSkbe8P6kD3p6Qt5BZAAgMDtdkOYiC4DlzYdrXFntt7YGtmi61Tt+LrSV/D9itbVDVWoYpThZUjVsrtnheXFgcwwNbMFgBkLtklvo300rtxaXEAB7Dtaqu2XiW68KYojy63jWgWZTRRhryMEL95/Pvvv/Hrr7/ihRdewOLFi9GlS5cWj6vMEJGWMll6CcaW8lveOVubg+oY5kIZTGShfNY/utCjQHp53GDPYFRUVuDh3YcYKhiKpUuXwsnJSWIf8aVnW9JSYVfR8+LvGTweT7kLIkRHtbgKDCGqEu86XFdXh3379mECm4Dznc7Ds7enzJBOLUwFv5qPyc6TYdXFqmnFlrRDGOs4Fvyapt4kwk8U+TV8lNaUSjwuPB+/mg9w1NttWRfeFOXR5bYRQtpfSxlx9epVHDt2DC4uLli0aBE6deqk1HGFGWuZaom+IX1l9oBoKZOlh5m0lN/yhqW0NgfVMcyFMpgQw6ALPQrECxC+7r4oKSlBzK8xeJ29Dr8AP/Tt21fhPoQQ5chdBYYYNnXMnixvlnxhGLtYuGDT/9uEHQ934O3Jb6PsizKcDToLAFiVuArxGfGi/aOSo5BwLwEu1i6InRMLXg4PCfcSkMXPQlxaXNOwl/8dO3Z2LFxsXJCQlYDN5zeLjsF14GLuoLlIKUxB1tOs1n9zpLS0WkB70uW2EUK0SzqTUwtTkdY3DYnXE2VmxMWLF3Hs2DG4ublh8eLFouKHMiugCHN++7+3Iy8vD+vD1gOQfG/hOnAROycWsbNlr7wlPIZwMj5F20pvLy4kJERUiFDlPU3e8VRBGUwIUYUwX4X3wP+9/F94/uSJFRErJO7Li4uLER0djYc1D5E9IBvlXcqVOq54/svLcXWsckWrsBB9RgWQDkq8225rCbsPRyVHNXuuuroaMTExOFF4Ane63EFSeVKz/cKTwkX7Sy/bJfza1dYVYGj6I0b4PBgk2hCeFI4bT24gPCm81ddFCCH6SDqTFWX0n3/+iRMnTmDQoEFYsGABTExM5B5HkcHzBsPI0giD5w0GoJ73ltZor/MSQogqpO+Bv/zzS9x4cgNRW6NEGVZYWIhdu3ZBIBCgyq0KR+4faTGPVcl/VTJeHspcos9oCEwHpY5uu+LFCvHJ5PqZ9UNsbCyePn2K9W+sx8nikxJdjGWNcZQm/GTuv5f/i0t5lzCqzyiZz6cWpsLW7PnYR/HZrIl20WSAhLQv6SEdwZ7BuJ1wG/u/249BYYMQEhICxhjOnj2LP//8Ex4eHpg1axaMjIwUHkeRbV9sQ5RPlOh9wG2OGxoONqj83qJoQlJZz0k/RkNRCCH6QPoe2NnaGdE3ojFm1Rj8FvkbPvzwQ+zatQvGxsZYtmwZChoLYGZm1mIey8p/8b8VbdsalLlErzEDMGLECI0e/+zZsxo9vi5pzbWmFKSw4duHM4uvLJjH/3mw4G+C2bjwcSwxOVH0/MqElSylIEXuMVYmrGR2X9uxlQkrJR4fvn0444Rx2PDtw1VuV0vodVUvR0dHBoA5Ojpq/FyKGNLrquls0xbKaPURXuuOHTuYo6Mj27Fjh8LtxX8uGxsbWWJiIgsNDWXx8fGssbFRrW2Tl+Ot3Vd4rbKea8u5dFFH/D9sCAwhozV5DYb0WrdEHdeam5vLvvrqK/bvf/+b8fl8iZxX5l66NVIKUpj/IX/mf9ifHU0/qtQ56HU1TIZ2rYqyjXqAkDaLSo5C1tMsNAgacKfoDopQhNoutdibsxeJBYm4V3IPp+6fwr2Se3Cxdmma4DT9kMRStrKq0amFqXC0cERtQy2ChgVhVeIqmZ8OEt1AnwYQoj3Krj4i/Llcv349jh8/jqtXr2L06NGYOnUqOJymdRPV1XurNZ8qCs/91sq3JIZBtnRcdXyCSQghuuLBgwfYu3cvzM3NERgYCEtLS4mcn9d3XptXrZJFfLWulMIU0fK7NLEqMWQ0BwhRirzJleIz4nHv6T0YwxgCgQA2sME/J/8TgcMCRfNzZPIzAQZk8jNF4x7j0uIQlxYnGn8oazK6qOQoXMy7iNddX8f9kvttHq9INIsmAyREezZs2ABHR0dRwVHehHQhISF48OABHBwccPXqVYwfP16i+AGobyx3ayYVFZ5753c75e4r67jqmMCUEEKUoeqEn/ImGZX3+HbedozdNRYPTB8gKCgIlpaWACRzXnquPHUJ9gzG/EHzMX/QfIR6hWrkHIToGuoBQpQinDAJaKoKC79OepCEtKI0PBM8Qw/0wL75++A1yAsARPNzCMc5iv8t7AGiKGQVfepHCCEdmfSSjfJ6hAgEAhw5cgQpKSnw9vaGl5eXRPEDaN/eW9RzjBCi65TtcSckfc+s6PG7d+8iPCkcT/AEl00vo3v37qLtpXNeE70yhCtwCfm6+6r9HIToGiqAEKXIm1RpqOVQ/Of0f8Ax5uC7+d/hZbeXRfsIh7aIT1QnDFbh3/EZ8fA/4o9Qr1CJ0JU16R11xyOEENlkFRKePXuGQ4cOIT09Ha+99homTJggc1/pm2xZNDXJsTLnJoSQ9qRqoVbeED3px9PS0nDw4EHMtpqNC50vIHwirWBIiDbQEBiiFOnuxlwHLv4x5B94kvQE/ub+OPnuSYnih1BLS23JW7ZWHUt0EUm0Zjshhkt6CFpDQwMOHDiA9PR0+Pj4yC1+SBMf3ijeVVv4CejHn3/crPu2rP0VbUMIIfpE1SG+8oboiT9+69YtHDx4EH369MG/3/k3brx7g3pfEKIlVAAhrfLw4UPExsaia9euCAoKgq2treg58Rtg6TGL8Rnx8PzJE/EZ8QCalq0d1nNYs2VrvZ280deiL7ydvLV2TYaO1mwnpGOor6/Hvn37kJWVhRkzZmD06NFK7yssPocnhUsUoTds2ABzO3MwL6awMK2N4jUVcwkh+iw5ORlHjhxBv3794O/vD1NT0/ZuEiEdChVAiMqys7Pxbey3OIIjuNTrEh7VP5J4XtENsHiPj9TCVPByeNg9e3ezqjcvh4fc8lzwcniavJQORXrSREKI4amrq8OePXuQk5ODWbNmYcSIESrtLyxaS0+GFxISgkuplxASHCJ6TFYhQlMT9YmjYi4hRF9duXIFv/76K1xcXLBkyRJ07ty5vZtESIdDBRCikszMTOzduxepnVNxW3AbR+8ebVboEL8Bli6GiPf4UFQo0cZNdEdDq7QQYthqamoQGxuLvLw8zJs3D0OHDlX5GMIu2r7uvi2uvCKrEKGN1VmomEsIaQt5vcgU9S6TNbyvpSF/0s/vObkHIcdDcMbiDLoO74q3fnsLAUcCmq2wSEMICdEsKoAQhcTD+NdLvyJwXyAemT9Cjxd7YIzjGDhZOcHZ2hkBRwIQcLgpxMVvgIM9g+Hj4gN+DR+phanwdfdF8jvJ8HX3VTjMxZCWOKTu2oQQTauqqkJMTAyePHmChQsXgsuVnZ3imS7MprAtYS3edMvKMXmFiNTCVAQcDpC4sVcnKuYSQtpCXi8yRb3LZH1o19KQP/Hnk5KS8P2F75FulI5r1dew8dxGxKXFIS4tTrQ/zX9HiHbQKjBEIWEY8/l8ZGZlIoWTguou1XiU8wh9LfqiuKYY0TeikV6cDjDA1sxWYrUWrgMXtma22HN7D2y7Sj4nPszFkCd+UnX5NHk0tQoDIUS/VVRUIDY2FiUlJVi8eDFcXV3lbivK9Bo+jnx+BJXFldiyaQtMVzeNQd86davEKlzCffaH7UfB4wKJHJO3gsvmc5ux7/Y+GBsbN8t9Qghpb/JWdVG02ouslV3EHytKK5K7z+D6wTh76SyWuC9B/879wQEHcwfNxaG0QwBH/kqLhBDNoB4gRKFgz2C83vN1NGQ1gNOVA99BvgifGC4aI+7j4gNHC0eM6TMGL1q9KLM3h/hwFkUTpKqDvBUMlNlHXZ9USh9PXd21O9K4d+o1Q4hyysrKEB0djbKyMvj5+SksfgDP8xgMYF4M5nbm+GTtJxJZLP4ppPDfg+cNhpWVFaqqqlr8uSytKwUDg2N3x2b5rq68VeY4HaU7OeUlIaqR14tMUe8yWT2TpR+T/lkcZD8I7tXuiEyOhD3XHq9OfBV2ZnZYM2ENfN19ETsnFrGzY8F14EoUnoXHUzXDpLdvzT2xouN1NJGRkbh9+zZlqwGiAghRqPpBNVzvuaLEogSZyISRsREOpR0Cv4YPVxtX2JrZ4mLeRZTVlSG/Kl80uSnwPDgBiN4gxG+sNTHMRd4KBsrso64uh9LHU1d37Y407r0jFXsIaa2SkhJER0ejqqoK/v7+cHJyanEfYe6umbAGIcFNE5uGfRImkcWWqZao/boWlqmWooLJti+2wdzcHCUlJS3+XFqZWqGzSWeMf2F8s3xXV94qc5yO0p2c8pIQ3SD+s8gYw7FjxxB9KxrpndKR1T0LO//eKTeTWjPEpqVjtOaeuKU2dSQRERGor6+nbDVANASGyHXx4kWcOHECbm5umO09G7tu7gK/mo+4tDiAA9h2tRV9uuft5I3wpHBkl2YjKjkKW6duFQUnAFEXaE137xNvDy+Hp9R51N0mTV2jvO7mhkhRN1RCCFBcXIzY2Fg0NDQgMDAQvXv3Vml/YSFElp3f7URlcSV2frdTVBwBlP+5XDN+jcT7gzh15aMyx+ko3ckpLwnRHFWGHwt/FtetW4f4+HjcuHEDIZ4hSOmcInPoTHxGPNaeXov+tv0x2Xlys3nxVM0weUNpVLknVnS8jmbDhg2ora2lbDVETMsSEhLYgAEDmIuLC/vqq6+aPb9nzx42ZMgQNmTIEDZ27Fh248aNFo85YsQITTRV5OzZsxo9vi4RXuuff/7JQkND2YEDB9izZ89Ez6cUpDD/Q/7M/7A/SylIkdg3pSCFrUxYKXpc+mtd0xFf147AkK5V09kmTRP5zBhltDqdPXuWFRQUsG+++YZ9/fXX7MmTJ2o/x44dO5ijoyPbsWOH2o+tio72unYUhnSt2s5oTdDkNRjCa+3o6MgAMEdHR4XbCa/12bNnLC4ujoWGhrKkpCSF+wzfPpwhDMw43JgN3z6c2X1tx1YmrGxVO1W95xbfXvjvo+lHlTqGIbyussh67zPUa5XF0K5VUbZpdQiMQCDAe++9h4SEBKSlpWH//v1IS0uT2ObFF19EUlISbt26hfXr1+Ptt9/WZhM7JOkxfn/88QfOnDkDDw8PzJs3Dxn8DInnbc1ssWb8mmZdm6WHtMgb4tLRxxQSooson/VDQ0MDdu3aBQ6Hg6CgIPTo0UOtx6fJlgnRXYmJiXBzc4Orqys2bdrU7Pm9e/fCw8MDHh4eGDduHG7evNkOrTQsqgw/fvbsGeLi4pCSkoIpU6bglVdeUbh9qFco3G3dMa3/NIR6hbZpXry2DJdp61AZQ0HDCTsOrQ6BuXLlClxdXeHs7AwAWLRoEeLj4zFo0CDRNuPGjRP9e8yYMcjLy9NmEzsk8aEqE+snIjk5GZ6enpgxYwY4HI7E8wCaDWtpy/lodQBCdAPls+7Ly8sDn89Hp06dEBgYCBsbG7WfQ12rVhFC1EtYpD516hQcHR0xatQozJw5UyKjhUVqa2trJCQk4O2338bly5fbsdX6T9nhx4wx/Pzzz8jKysK0adPw0ksvtbiPr7uvxCqIbVkRsa3DZYDWD5UxFDScsOPQag+QR48eoW/fvqKvHR0d8ejRI7nb79y5Ez4+PtpoWocmnOSuf2V/lFaV4mbPm3Ae7QwOhyPxfLBnsNyVW1Tp1SF9jNTCVAQcCUDA4QDqFWJgaHUC/UH5rNsePHiA2NhYGBkZISgoSK3FD/GfU+GnnW+tfKvNq6xoeoUtQjoS8SJ1586dRUVqcePGjYO1tTUAKlJrU319PZ4+fYp79+5h5syZShU/1E3VhQXEtxf+29fdV+2LE+gTdS1aQHSfVnuAMMaaPSb8JVva2bNnsXPnTpw7d07m8zt27MCOHTsANH0qxuPx1NZOaZWVlRo9vi7wrvNGdUM1Opl2greRN25cvoFii2LR876mvqI1zoX/5qXxRM/nlefBqcYJf1/6G0UWzddClyZ+jLzyPAyrGwYwKL2/OnSE11Wova61trYWK1euRG1trdbO35FeV3VSZz4DlNHqJLy57t+/P0xNTXHjxg21Hl/853TIkCHYvXs38srzwC/jK8zklnJf1fcFadKva1uPp8sM/f+wuI50reokq0itqHcHFanVK7UwFZvPbQY4kBgGXltbi29ivoG9hT2GTBoCT09PiSVtL8RfoGGFhOgYrRZAHB0dkZubK/o6Ly9P5sz1t27dQnBwMBISEmBrayvzWG+//bZo/PnIkSPh7e2tkTYDAI/H0+jx24NwnPe6detgZ2eH27dvw9vbG7XPanGm6AxSslNg19UOxTXFGOwwGF9N+kphRTi1MBWfnfkMWQ+z8NVrX6nUjU+4byY/E5uGb4K3u7dq+57+DFklWfhqkmrnNcTXVZ72utbMzEx8/fXX2LBhg9bO35FeV3VSZz4DlNHqcuzqMWxM2ggPCw+YuJpgetfpMq81MjIS68PWY/C8wdj2xTaZcy8Jb8iln5P1cyrc/jXP1+Rmf0vbKHMMRYSvq/D96q2Vb6GMW6bU8fRtLhND/j8srSNdqzrp44eI+lbsKi4uRn5+Pnr16gU7OzuJ5zKKMzC0YSg44IiKsIwx8Pl82FvYw8bUBtVPq8Hj8SSKte3xQZBQ7bNaFFcXw87MDqYmpiirLcPjysfobd4blqaWcrdriT69ropeU2Xo07W2VUe6Vq2uAtPQ0MBefPFFdv/+fVZXV8c8PDxYSorkTMMPHjxgLi4u7Pz580ofl1YYUJ1wVmsbGxsWGhrK/vrrL8YYYwmnEpj91/YMYRD9MQ43Zv6H/GXODC0+i/Tw7cMZJ4zDhm8frnJ7ViasbNXs1ysTVjLjMGOGMKh8XkN8XeWha9VP2lxhQFP5zBhldGulpqayMWFjmHm4ORv6w1Bm97Ud2x2/W+a2VlZWDACDKWTmaGsztj0JX1dlV2EQ15p92pOh/h+WxZCuVZsZfeHCBTZlyhTR1//617/Yv/71r2bb3bx5kzk7O7M7d+4odVxaBeY5RbnRJaILQxiYUZgRSylIYVdyrrBXN7/K3o94n/1+5XcWfSSa+R9qWiHxaPpRNnz7cHY0/SjbsWMHs7KyYtbW1mzHjh1aXSFROvfl3aer+v6gT69rW98L9Ola28rQrlVnVoExMTHB999/j9dffx0DBw7EggULwOVysX37dmzfvh1A0wRsfD4fK1aswLBhwzBy5EhtNrHD+OKLL2BjY4Nx48bBx8cHEyZMAAAUVxej9lktuhh1gTGMYdHZAtP6TwM4kDkztPgs0qFeoRjWcxhCvUJVbo+8uUWU2W9a/2kYaDewVeclhDShfNYtt27dwsGDB+HTwwfLRixD+MRw+A3xg52Z7E+whJ8EdzHpIjNHW5uxukCVVRjasg8humzUqFHIzMxEdnY26uvr8fPPP2PmzJkS2zx8+BBz5szB7t27MWDAgHZqqf5SlBsfj/0YpsamWDN+DV7o+gLW/LwGl2ouocqtCm+MegMmRiZIuJeAqOQo8HJ4yC3PBS+Hh5CQEJibm6OkpAQREREqr9bSFtK5L+8+XZ/fH1pC7wVEJi0WYjSGPl1UTV1dHYuJiWFhYWHs2rVrosdTClJY9JFoNmPvDOYd7c0G/ncgO5p+VPSc/2F/UXVbfB9tVbLVzdBeV0XoWvWTNj9d1CTKaNVcv36dhYWFsejoaFZXVyfxnLxr3bFjB3N0dGQ7duxQeGxlt9MFhva6KkLXqp+0ndHHjh1j/fv3Z87Ozuyf//wnY4yxH3/8kf3444+MMcbeeustZmVlxYYOHcqGDh2qVPuoB4hqSktL2bZt29iH//yQvfXLW6L734RTCaL7Yel7Y/Hcbe19s/h+8v4ta1tN0Obr2t7vWYb4f1geQ7tWRdmm1TlASPurq6vDvn378PDhQwx+dTD2FO2BaWHTeL/5cfMRbBOMlKIUPCh9AAaGz05/BlcbV3AduLDtaos9t/fA1swWW6dubZoQ6vxmoPmwVEIIIa1w5coVHD9+HK6urli4cCE6deqk1H4tLdUonBOjsrISpaWlePfdd0X7EUL0w7Rp0zBt2jSJx5YvXy76d1RUFKKiNN+zoKN6+vQpYmNjUVtbi08DP4Wjo6PoOVMTU2x9bavo661Tn/9bOp/Fn1OWsOeIkKx/C48rvm1rzqVLaGl2oglUAOlAampqsGfPHuTn52PevHmIypUM0/sl98FsmKigwcBwv/Q+opKjsHXq1mZrhkclRyEuLQ5gEBVFCCGEtM6FCxdw8uRJuLu7Y968eTAxUd9btPAm0traGsbGxhAIBHRDSQghSiouLkZMTAwEAgECAwPRq1cvrZ5f+h5cmX8bwpCWDRs2iCa0JkRdtDoHCGk/VVVViImJwZMnT7Bw4UJwuVwEewbDx9UH/Go+vJ28sYC7ALZdmwoZS4csxYwBMzDFZQr4NXykFqY2O2awZzAmO0+Gk5UTvJ28ATTN+r8qcRXiM+KxKnGVzP2E26QWpkr8mxBCOiLGGJKSknDy5ElwuVzMnz9frcUP4Pk46M2bN+PHH3+EtbU1KisrERkZqdbzaEJkZCT69u2rF20lhOi/yMhIWFtbw8bGBpGRkUhKT8KUH6dgX80+9PDqgXd472DQ94Pw+enPMej7QZi5fybKastE978BhwMQcCRA6XtbZe6FuQ5cbJ26FVwHrsx/X4i/IMpJ8ef1XUhICHJzc6lYT9SKeoB0AJWVlYiJiUHG0wzk9ctDZmom5rK54OXwAAAJ9xJga2aLNePX4O9Lf8PVxhVrJqzB5nObcS3/GvIr82HbtWm5yz2394BfzYetmS2CPYPhYu2Ci3kXwcvhwdfdV9TtLjErETmlOeBX8xE7J1aiPfK68VEPEkJIR8MYwx9//IG//voLQ4cOha+vL4yM1P/ZhHQXbGW6FWtyKVlVjk1doAkh2hQREYHS0lIAQGhoKJxXOON2422YGJtg241tuFVwCwImQPalbNQKanH36V34DPLBnuw9SHqQhPSidIAD0YeKLVHHkBXKSUKURwUQA1dWVobY2FhUVFSg2q0aiZmJAANSilKQW54LHxcf0czPUclRcKpxEs1MHZcWBwaGQfaDJLrR8Wv4oqCW7mYn/PteyT3klOUAMpaob6kbHyGEdASMMZw4cQKXLl3CiBEjMH36dNFqLpqmTLdiTd5Qq3Js6gJNCNGmDRs2YPXq1WCMYcyYMejftT/69OmDLl26YO7Audj5905k8bMwy30WjmYchautK3qb94bfED94O3njUNohgKP8va06hqxQThKiPCqAGLCSkhLExMSgpqYG/v7+KO9SjoZODQAD5g5q6gES7Bks6iIX7BmMvy/9jdc8XwMA8Kv5AAdYM36NaBvh5Ke2XW1F+4pXq4VfpxamwsXaRWaYS+9DPT+e0+QnroQQ3cEYw7Fjx3Dt2jWMHj0aU6dO1XjxI7UwFVHJUQj2DJY5aar481wHrswbaultVDmn+Paq3Ky3NMErIYSoS2RkJNaHrYfTbCegH+Bm4QZfX18cuHtAlGO+7r6ibItbEAeuAxc8Hg9bvZvuZ33dfVU6p/R9cWtQThKiPJoDxEDx+XxER0ejrq4OgYGB6Nu3L7gOXMTOjkXsnFj4uvs2Gx/IdeDCzsxO1AMkdk4sYmfHNrvJ5TpwRT1GFM31QVQn/qkoIcQwNTY2Ij4+HteuXcOECRO0UvwAnnezFmZ8S8/LGnvd0jGUPWdISIioCEJzexBCNEnRPELxGfEY9H+DMHPfTKQWpmJ92HoUPC7AjUM3kIIUlLuW48DdA81yTNUsJIToDiqAGKDCwkJER0dDIBBg2bJl6N27t9L7FlcXKxXowuAPTwqXuT29MbSOcKLCtnZhpEkDCdFNAoEAhw8fxo0bNzBx4kRMmjRJKz0/ViWugreTt2jIoyzBnsEynxefoE/eNvLOy6/hw8fFR+b2igq+qYWpyCvPo+I6IaTNZGWNMNc+O/0Z0ovTcSzzGPyP+KPnlJ6ABdDduzvmDZyHFaNXNC0a4OIDfg0fYVvC0LNPT/xx8A/Ym9nD28m7VXmVWpiKgCMBCDis/GSpuqK4uJjuMYleowKIARC/Oc3Pz8euXbvA4XAQFBSEYk5xsx4aqYWpEjNUi+9vZ2YnurlVtKKL8CY4aFgQ+lr0Fa0CI/08ze2hGnXNdk09SQjRPc+ePUNcXBxSUlIwefJkeHl5qb34Iav4KSxI83J4ClcGkLdywEcbP8J/Fv0HH238SKXVBaKSo5CQ1TTJtqztFRV8o5KjwK/hUxGdENJmsrJGmIuutq4YaDcQ/Sz7Ias4C8yJYdL6STj7/VnsX7BftNKKrZktErISsGXTFhQ8LsCtuFvILs0GL4fXqryKSo5CXFoc4tLi9C7n8vPz6R6T6DWaA8QACEO8srISLlku6NKlCwIDA2FjY4ONiRubzSwtDF3hDNXA85VYfE19sfW1pu1WJa7CnttNM1rnludKHEN4E7wqcRVyy3NFq8AIqWM8I2k9mgyLEN3S0NCAAwcOICsrC9OmTcNLL72kkfPImly0rRPspRxMQWNZI1IOpgDblN+vpfMqGrMuPScVIYS0lqysEc8nrgMXh88dxjdnvoFDpgOu864j2TIZpr6mojmMgj2Dca/kHq6+dhUmf5jAfY47+g/qLzrO1QtXwa/mI7UwVWbBV3o+pGDPYPBr+ADTzEIAqs7XpMoxevXq1aygpI7ztTeah6/joAKIAQj2DEZFRQXM7pjBrLsZAgICYGVlJXpO/G/hv4UTnEo/XpRWJPE1AHg7eYsmTJV1bunjk/ZHk2ERojvq6+uxf/9+5OTkYObMmfD09NTYuWQVP9takP4y7MtWFVTbcl6uAxdFFkV6eyNNCNFt4vl0/fp13D5zG8udluOLyC/w6NEjrA9bj96NvZFdmg2g6QPAvPI8FLgXYJj3MFx85yKApjlEwpPCscJ+BRLuNfV4E86TJ14MEF/q1tvJG+FJ4Qj1CpX48FDZIoIy26ljaV15x7Czs0Nubm6bzicsNnh7e4PH4+lE0YGWEu44qABiALpWdoXZHTPcMrmFpTOWioofgOwbUK4DF7FzYgE0D1FeGq/Z8V1tXFWe0ZoQQghQW1uLffv2ITc3F7Nnz4aHh0ebjiccey3vZlETxU8qqBJCDNXly5eRkJCA/v37Y8GCBaivr0dERATc5rjheul1vGj1ouhDvlCvUFHhQig8KRw3ntxAnU2daOi3rGKA+AeG/kf8cePJDYQnhUvcXytbRFBmO3V8QKnKMVQ9n7DYsH//fggEAp0oOlDv6Y6D5gDRc3fv3sW+ffuQ2jkVt41uY/+d/Srtr2iyUmUmMqXJTgkhRLaamhrExsYiLy8P8+fPb3PxA6Cx14QQoi67T+zG2wlv4zfz35BgmoC7JXcREhKCxOuJ6O3VGzMGzMDu2bsBNA0LBwCvfl5wtXEVHSNoWBDszOxgb2Yvmh9Jeh68sC1hGMMdA8tUS3AduAj1CsWwnsMQ6hUqMW+TsvPnKbOdKvM1qeMYqp5POC/L4sWL1TL5vzqoax4+ovuoB4geS09Px8GDB9GjRw987fM19qTuUbnSq6hi21I1NzIyEvvD9mPovKEIDqQhMIQQIlRVVYXY2FgUFxdj0aJFGDBggMLtUwtTsfncZpTWlcKqixXWTFgj80ZS1tjrlo6r7+OyCSFEnRhjSEpKwv9d/D+kG6WjsboRaelpsO/WVMSISo5Cwr0E+A3xA9eBq3BOvPsl98HAUC+oFx1fuvf1lk1bUFlciS2btiDskzD4uvuKen70ndxXVNTODcnF1qlbRSvEgEHme4Gi4YWazHzhsV83eb3Nx2qvnoU0zwcBqAeI3rp9+zbi4uLQu3dvBAQEwLSLaauOI6xURyVHiVZ5Ea7+AkBUzRVfKUYoIiICBY8LcOfwHbqxJoSQ/6moqMCuXbvw9OlTLFmypMXiB/B8curjmcflrgoQGRmJ/Px8lW7cqJceIYQ8xxjDmTNnwOPxsNR9KV53fR39LPthsvNk0QqI/Go+fFx9JD4I9Bvih1CvUImeF6mFqbhXcg92Xe3QvXN3uef8ZO0nMLczxydrP2n2nLwValRZIUa8F4kmM1947OLqYrUfW1tolUQCUAFEL/399984fPgwjHsa42avm7hXfk9m4EmvMS6riAE0v0GWdSxZjylawpAQQjqisrIyREdHo6ysDEuXLoWLi4tS+wV7BmP+oPmY1n8a5g+aL7PnXUREhGh8urJ0dUlyWcv1EkKIJjHGkJiYiHPnzmHUqFF4f+H7cLVxRXl9OVxsXMB14Ip6f9h2fb58t7DHha+7r8Qwj6jkKJy6fwo5pTmoqK+Qe96wT8JQUVSBsE/Cmj0na9iF8P1A3nuBNPFf6jWZ+cJj25nZqf3Y2kK/uxCAhsDonStXruD48eNwdXXFVZur2J+6H6W1pQAH8HHxkQg80XK3DLA1k1zuVrzrnHiFuyitSOJrYXc3Z2tn9LXoC28nb9F+ISEhGOc7TtR7hHqBEEI6sqdPnyI2Nha1tbUICAiAo6Oj0vuKT04tz4YNG1BbW6vSjZuuLkkuPds+dUsmhGiCcJWWMU/G4Jf//oKxY8di+pvTcdr4NJyKnODt5I2kB0mi+1tVJ/4ULmWrzqIA14GL2NmK3w/EiU/eqcnMFx6bx+Np5PjaQJN6E4AKIHrlwoULOHnyJNzd3TFv3jwMfzocRkZG4NfwkZD1fKyikLw1xqVDXTwseWk8ia+F4x77WvRFbnkueDm8Vs1YTQghhqy4uBgxMTEQCAQIDAxEr1691H6OkJAQ8Hg8eHt7q/3Y2iY92z4tP0gI0QThKi23tt6CoEyAy5cvw/U9V9G9KwCJ+1tVCgjihYr2LArQL/WEqIYKIHrizz//xB9//AEul4s5c+bA2NhYFNKphamw7Wors7AhXUFWtUghPKa3kzd4Obxm51DHMluEEKLPCgoKEBvblLXLli2Dg4NDO7dI90nfsNPyg4QQTVj38jr849d/wPZlWzy4/AAbN27EOM9xABR/OEgIMVw0B4iOE07W9Mcff2Do0KGYO3cujI2NJbaRt/SUrDk/5M0DIlT7rFbieeGxxZf8UubchBDSETx+/Bi7du2CsbExgoKCNFb8EM6ZUVysv5PPKULLDxJC1O3Zs2eou1mHgNoA/Ofz/6CwsBAhISES9650H0tIx0MFEB3GGMPJkyfx119/YcSIEZg1axaMjJR/yZSdzFRccXWxzOdpJQFCCJGUm5uL2NhYdOnSBUFBQbCz09zEcMIhIvn5+Ro7ByGE6BvpD/aEXx9KOQTv77yx6c4mVAyqwNIrS9H3333hFuiGbnbdsCJiBVYlrkJ8RrzcDwbjM+Ix6P8GYea+mc2OL2v7yMhIWFtbw8bGRuUJnmliaEK0h4bA6CjGGI4dO4Zr165h9OjRmDp1KjgcjkrHkDU8paUhK3ZmdjJnj6ahLrqPJhEkRHtycnKwb98+mJubIzAwEJaWlho9n3CIiHBukdTCVGw+txngAGvGr1HLp5fCSa+DPYPlHk+ZbQghRFvE56IL9gyG/xF/ZD3NQtSVKNSwGhgZG+HO3TuoE9Q17XAUQDkQuTUSVp9b4UTWCWSXZoNfzUfsnFhERkZifdh6DJ43GI8HPEZ6cTruFt+Fi40Ltk7dqnDuu4iICJSWlgIA3n33XQBQ+n6M5kEiRHuoAKKDGhsb8euvv+LGjRuYMGECJk2apHLxA5A9+39LkzuZmphi62vNn9fVlQTIc9Jvno2NjSguLkZpaSksLS2Rnp7e3k3UCn26VlNTUzg6OqJTp07t3RSignv37uHnn3+GlZUVAgIC0L17d42fUzhnhnCiPdEqXxzAtqutWvJZmUmtaeJr9aGM1m3GxsawsrKCnZ2dSr1viXaJf0C3+fxmpBalwkRggnrUw9TYFMbGxhJL1HZ+tTPq/6jHoLmD8OqQV3Gv5B4yT2fi0HeH8DL/ZURERKDgcQGKYorwxg9vAABcrV2bfRAo6wPBDRs2YPXq1SgrK4NAIJAoZogXjwE0KyS/tfItbNm0BW+tfEvmdXaE4rMufZAnzGcLCwu9yCt10JdsBtqez1QA0TECgQBHjhxBSkoKJk6ciFdeeaVVxQ/S8UhPIpiXlwcOhwMnJyfU1tbCwsKinVuoHRUVFVr5hbStGGPg8/nIy8vDiy++2N7NIUq6c+cOfvnlF9jb28Pf3x/dunVrl3YEewaDX80HOOrrmadMTz/qDag+lNG6izGGhoYGFBQUIC8vDy+88EJ7N4nIIf4B3bNnz9AoaERf9IWRlREyyzMhqBeAgQEALDpZ4J9r/4n779yHt5M3/vvjf3H95+torGhEdU01Plz7Iaxet0LX37tC8EwAxyxHuLzkIlF04DpwEewZLLMYISxUi/8iLyRePAaAyKhIRCVF4d//+jdCQkJQxi2D6WpTlHHLZF5nRyg+61IvGPF8tra27hC/i+lDNgPqyWcqgOiQZ8+e4eDBg8jIyMDkyZMxfvz49m4S0SPSqypUVVXBzc0NRkZGqKura8eWEVk4HA5sbW1RVFTU3k0hSkpNTcWhQ4fQq1cv+Pn5oWvXru3WFq4DF7FzYlveUMVjtnRjTb0B1YcyWndxOBx07twZffr0wZ07d9q7OUQJlZWVcHniglEYhXXT1qFfv36Y/8t8ZPAzRNvUNtbifsl9bJ26FasSV+FszFk0ljWC05WDTtadUDe+Dk/cnqBTQic0VDRg9393w3S1KQDJokNLxQhZy9JKF4+jkqJQWVwp+mW/peJyRyg+69JqYMJ8rqqq6hDFD32ijnymPn06oqGhAQcOHEBGRgamTZvWIYofNOGT5lG3Xd1Gb6r649atWzh48CAcHR0REBCgcvFDOHGerAn3WlqdqzXnUcexiOZRRus2en30Q3l5OXbt2oWysjK4D3RHv379wHXgIm5BHKYPmA7H7o5wNHfE6y6vSxQSXH1dwbHkoP+C/oi7EIfpi6ZjoN1ABK8KhqOjIz5Z+4ncefH8hvjB28lblLfSqyhKEy8ef7TxI3AaOLC2thb9st/SajQX4i/gYMhBXIi/oK5vm87RtdXA6Odft7Xl9aEeIDqgvr4e+/fvR05ODmbOnAlPT8/2bpJW6FJXN0IIkef69ev4/fff4eTkhMWLF6Nz584qH0P4iWHSgyTklucCeP7JoTq7NneEbtKEECJUWlqKmJgYVFdXo9qtGr/l/AbLZEtRMeHXxb9iVeIq7Lm9By7WLhJDWUoGl4A5M5SYlcDX3Re+7r6i4/6w4Qe55xQWK4THBYAR9SOw507L2RuVHCXqeeLo6EiTpBLSDqi01c7q6uqwZ88e5OTkYPbs2R2m+AE0dXVzdHTUia5u6qAvPVqKiopw69YtGnpBiBIuX76M3377Da6urliyZEmrih+phangV/Ph4+qDoGFB6GvRF95O3qLnhZ8mSn/KKOzN8d/L/4XnT56Iz4hv8VzBnsHwcfEBv4aP+Ix4BBwJQMDhAOoRQggxOE+fPsVXUV8hpiwG11+4jqEvDEVfi75wtnZGwJEAzNw/EwGHA+Bs7dwsd1MLU2GdYg3OVg4mPZ0kcVzxnnSphal47aPX0LNPz2b3d95O3qLjyltFUVqwZzAmBk5Ej949VLr/3bBhA3r07gG3OW4az3N9uZ8lpLWoANKOampqEBsbi7y8PMyfPx8eHh7t3SSt0rWubm0lXp3XZfn5+aivr0d+fn57N0Vrli1bhunTp6u0j7e3N95//30NtYjog/PnzyMhIQHu7u5YuHBhq1friUqOQsK9BNh2tcX9kvvILc8FL4cnel5e12dhb44v//wSN57cQHhSeIvn4jpwYWtmi4SsBIQnhSMuLQ5xaXGISo5qVdsJ0QbKaKKqoqIiREdH40LtBaQZpeFYzjFE34hGbnkuom9EIy4tDsczjyMuLU70uHjuRiVHISs+C6yM4dyecxLHFmZvVHKUqMdGweOCZvd3vBye6LimJqYKh7AIcR24OL3tNJ48eqLS/W9ISAgW71yMm443NZ7n+nI/S7TDEPOZhsC0k6qqKuzevRtFRUVYuHAh3NzclNpP2WWwVFkuS3pprtYep6PTpcmbFOnVqxfy8/PRq1ev9m5KMy3NiREYGIhdu3apfNxt27aBMabSPocPH6blaTsoxhiSkpLA4/EwePBgzJ49G8bGxq0+nqzJ65SZyE64jbO1M6JvRCPUKxR4ovz5vJ28cSj9EMBaP3GeLi1LSNofZTTRBU+ePMHu3bvB4XCwcd5GRKVFARxg7sC54OXwRNlXWlsKqy5WmDuo6XHpDL4deBspB1Oa3bdJZ7Yy2xWlab5XrbYmQtWX+1kiifJZeVQAaQcVFRWIjY1FaWkplixZAhcXF6X3VXZ8tyrjwMW39TX1lfscjSdXTNas37rI3t4e9vb27d0MmcR7pfz+++8ICQmReEx64smGhgalAtbS0lLlttjY2Ki8D9F/jDGcOXMG586dw7BhwzBz5sw2T4QmvXKKslkqvt8Hoz8AAPCe8FTaT3xMe2vQuHMijjKatLfHjx9j9+7d6NSpEwIDA2Fra4tX3F8RPS/MPOnsk/5a2BMD25qfQzqzldmOl8Zr5RUpT1urcOnL/SyRRPmsPBoCo2VlZWWIjo5GWVkZli5dKrf4IW8mf+mx4spup4j4tmW1ZRJjzVU5DiFt1bNnT9EfKysricdqa2thZWWF/fv349VXX0XXrl3x008/gc/nY/HixXB0dETXrl3x0ksvITo6WuK40t33vL29sWLFCnz++eews7ODg4MD/vGPf6CxsVFiG/Hue05OTvjnP/+Jd955BxYWFnB0dMQ333wjcZ67d+/Cy8sLpqamcHNzw/Hjx2Fubt6qijvRPsYYEhMTce7cOYwaNQq+vr4dfhZ4dc3VRGPKDQNlNGkvqYWpCDkYgjX/bw0SGhNQP7wekw9OVmpuJGmUR8QQtTWfHRwcwOVyO0Q+d+w7Oy0rKSlBdHQ0qqur4e/vDycnJ7nbio8/FCc9VlzZ7RQR3/Zx5WOJseaqHIcQbfjss8+wYsUKpKWlYdasWaitrYWnpyd+//13pKam4t1338U777yDM2fOKDzO3r17YWJiggsXLuD777/Hd999hwMHDijcZ+vWrRgyZAiSk5OxZs0arF69GhcvXgQANDY2Yvbs2TAxMcGlS5ewa9cuhIeHo66uTm3XTjSHMYbff/8dly9fxtixYzFt2jRaphjqm6uJxpR3HJTRRBP+zfs39qXuw6nGU7j47CLWfbcOf6/+Gx/88wOF+8n6oFA6j2j5cNJRKMrny5cv46OPPuoQ+UxDYLSkuLgYMTExEAgECAwMbHH+BWXH+al7PGBv894Y1nNY01hzQnTQBx98gHnz5kk89umnn4r+HRQUhAsXLmD//v2YNGmS9O4igwYNEt38DBgwAJGRkThz5gwWL14sd58pU6aIKtoffPAB/vOf/+DMmTMYO3YsTp06hTt37uDkyZPo06cPgKawHz9+fKuvlWhHY2Mj4uPjcfPmTbz88st49dVXqfihZjSmvOOgjCbqlpWVBbM7ZhhtOhqmvU1xMvskBDwBUA5Un6lWuK+sodzSeUTDvUlHoSifKyoq4OHhgT/++MPg85l6gGhBQUEBoqOj0djYiGXLlik1+aSyPS8UbdeairalqSWS30lu87jx1rZB3VV4quqrhy59H0eOHCnxtUAgwMaNG+Hh4QFbW1v06tULhw8fxsOHDxUeR3rVpd69e6OwsLDV+2RkZKB3796i4AaAUaNGdfghFLpOIBDg0KFDuHnzJl599VVMmjSJih8aoOurfulzl3hdymeAMpqoT2phKpbtX4Zt+7bBxtoGXC4XPgN8MMBmADzme6BH7x4YO2Esutt3R9iWMMRnxIuGcQt/LrydvJsN5Q4JCUHi9USk9U1D2JYw7H9rP4bmDaXh3kQjdCmjFeVzv379YG5u3iHymVJfw/Lz87Fr1y4YGxsjKCgIDg4OzbZR9gcjtTAVM/fPxMDvByI+I14i6GWRNzympfOp8wdVXhvUtb22j9dR6dL3sVu3bhJff/vtt9iyZQs+/fRT0eSVs2bNQn19vcLjSE/8xOFwJMYvqroPY4x+cdYzz549wy+//ILU1FRMmTIFr7zySss76Qht/8KuzwUCZejzEB1dymeAMpqoz+bTmxF3Nw4ZZhl46vQUP6f/jB+u/oDs0mz0m9QPi3cuxh9n/0BlcSW2bNqC8KRw3HhyAyG/hWDzuc3Yc3sPeDk8iQ8Khfe4wue3bNqCgscFuHP4jswPEw09+4jm6VJGK8rn3377DTdu3OgQ+UwFEA3Ky8tDTEwMunTpgqCgINjZ2cncTtkfjKjkKBzPPI4MfgbCk8JFQS+cr0NIGO7O1s7oa9EX3k7eEo9vPr9Z4flktae1RRFVJ1Ft7aSr8t6gaBJX9dDl7+O5c+cwY8YM+Pv7Y9iwYXB2dsbdu3e13o6BAwfi0aNHePz4seixa9eutfiGQNpHQ0MD9u/fjzt37uCNN97AuHHjNHo+6QyVl6nKZq22f2GXdb7WvC/o6i8T6prstT3ocj4DlNGkdW7evInumd0x3nw8vl70NZa/tBx+Q/zQ37Y/wAGy+FnYc3sPPOZ7wNzOHJ+s/QShXqGwM7ND7bNagAOZPxei4S7/e/6TtZ8o/NnX5+Io0Q26nNHi+ezh4QEXF5cOkc9UANGQBw8eIDY2FmZmZggKCoK1tbXcbZX9wQj2DMa0/tPgbuuOUK9QhHqFypyvQxju0TeikVueC14OT+JxMNlvCora09rqpaqTqLZ20lV5b1A0iat66PL3ccCAAaJPFTMyMvDJJ58gOztb6+2YPHky3NzcEBgYiJs3b+LSpUv4+OOPYWJiQp866pj6+nrs3bsX9+/fh6+vL0aNGqXxc0pnqLxMVTZrtf0Lu6zzteZ9QVd/mdD1ITqK6HI+A5TRRHXXr1/HT0d+QpZFFr5e9DWGOw4XPffW8LewfMRyfPXaV/Ab4oeo8ChUFFUg7JMw+Lr74mzgWbw1/C3MHTgXtxNuY9LwSRIFV+E97prxa7B16laEfRImGg4jq5irz8VRoht0OaPF8/nu3bt4//33O0Q+0ySoGnDv3j38/PPPsLKyQkBAALp3765we2XX9eY6cPHr4l8lHpM1V4ewcOHt5A1eDq/ZRKnBnsEKfwhltUfdk62qG02w13GtW7cO2dnZ8PHxQdeuXbFkyRIsXboUaWlpWm2HkZERjhw5guDgYLz00ktwcnLCli1bMGfOHJiammq1LUS+2tpa7N27F48ePcKcOXMwZMgQrZxXVg6L/y1vO3lCQkK0+su6rPO15n2BsrrjoYwmqrh06RISExORbZ2N5LpkxNyOwbA+w2ROVCrrHlh4D7sqcRXOxpxFY1kjIiIiRPkl6x5X0SSo2s5aQrRJPJ9NTU0RFBTUMfKZGYARI0Zo9Phnz55Vets7d+6wiIgI9uOPP7LKykrNNUpDVLlWcTt27GCOjo5sx44d6m2QBrX2WvVFWlqa6N/l5eXt2BLt0qVrvXHjBgPArl27Jncb8ddJmqazTVt0JaOrqqrYTz/9xCIiIhR+39sqpSCFrUxYyY6mH2UrE1aylIKUZs+JP6boGNLbGXpuiTP0a6WMbn+U0Zq9BlV/hv/66y8WGhrKDhw4wG7m35TIQHmZeDT9KBu+fTg7mn5U4vGUghQ26cNJrEfvHhL3prKOI/2YsjmtyrXq432yPIaezYw9/7nXpbzSNF27Vk3mM/UAUaO0tDQcPHgQvXr1gp+fH7p27dreTdIa8W7NVCknHdWRI0fQrVs39O/fHzk5Ofj4448xdOhQeHp6tnfTOryqqirExsaCz+dj4cKFGDBggMbOJfw0MelBEnLLcwE8/1RR2eUWaVlGQtSPMlo3McbA4/GQlJSEIUOGYPbs2TAyMpLIPnm9pcXnwxPvEcJ14OL0ttPANsntZWWr9LE1kb90n0yIYtrMZyqAqMmtW7dw5MgR9O3bF0uXLkWXLl3au0laRd2aCWlaQ33NmjXIzc2FtbU1vL29sXXrVhpf3s7Ky8sRGxuLsrIyLFmyBM7Ozho9n7xhiOLPKTPnkzLbEUKURxmtexhjOH36NM6fPw9PT09Mnz5dpaUvQ71CEZ4U3mw+PHmUyVZN5C/dJxOimDbzmSZBVYPk5GQcOXIETk5O8PPza3Xxo62rAqi6vTqXu9XlCeR0ddUBYngCAgJw9+5d1NTU4PHjx9i3bx969OjR3s3q0EpLSxEdHY3y8nL4+fm1qfghzMz4jHiF2Sn8NNHX3bfZxGeqTIbGr+Fj87nNaslodaNcJfqIMlq3MMaQkJCA8+fP46WXXsKMGTNUKn4ATfOAJL+TLHM+EFmUyWBNTFqpy/fJhOgCbeYzFUDa6MqVK/j111/h6uqKJUuWoHPnzq0+VltXBVB1e11al1qTdHXVAUKIZj19+hTR0dGoqalBQEAA+vXr16bjCTMzPClc49kZlRyFuLQ4xKXF6WRGU64SQtqisbERv/32G65cuYJx48bBx8eHeuIQQrSChsC0wfnz53Hq1Cm4u7tj3rx5MDFp27ezrasCqLp9R+liTd0OCel4ioqKEBsbC4FAgMDAQPTq1avNx1Q0tEXdgj2Dwa/hA0w3M5pylRDSWo2NjTh69Chu3boFLy8veHt7U/GDEKI11AOkFRhjSEpKwqlTpzB48GDMnz+/zcUPQH6XO1W74im7varH1dcuz9TtkJCO5cmTJ9i1axcYY1i2bJnC4ocqQwEVDW1Rx/GlzxU7Oxaxc2LBdeAitTAVAYcDkFOaoxNDYihXCSGtIRAIcPDgQdy6dQuTJk3CxIkT5RY/xPOzrcO2UwtTEXAkAAGHA9qcofp6P0wIaUIFEBUxxnDmzBmcPXsWw4YNw5w5c2BsbNzifuqa30PV46tTa7o8a6NdhBAi9PjxY8TExMDY2BhBQUFwcHBQuL2mhwKq6/jCITFPa58qfSy6SSeE6JJnz57hwIEDSEtLw9SpU/Hyyy8r3F48P9uapeocVkhDAAnRb1QAUQFjDCdOnMC5c+cwcuRI+Pr6Kj1ZU1vm91BUARd+vfncZpWO0xobNmyAo6OjSl2eO8o8I9LoFw9CtC83NxcxMTHo0qULgoKCYGtr2+I+wZ7B8Bvip7FhJqocX1FGB3sGY/6g+bAxtZE4lqJ9VLlJTy1MxWsfvYaefXoiMjKSiteEELVqaGjA/v37cffuXUyfPh1jxoxpcR/x/GxrVgszdP6g+W3O+9bcDxNCdIfWCyCJiYlwc3ODq6srNm3a1Ox5xhg+/PBDuLq6wsPDA8nJydpuokyMMRw7dgyXLl3CmDFj8MYbb6g0XlFecCsT6Ioq4KK1yjlQ6TjShL+wFxcXy92/NV2eNf3LhS6QVeygTweIPtLXfAaAnJwc7N69G+bm5njzzTdhbW2t1H5tme1fXkG6pVVi5IlKjsKum7vgf8S/2b5cBy5i58TCycoJAETH33x+M7Zf347N5zc3O54qN+lRyVE4G3MWBY8LEBER0ez9ggoihJDWqqurw549e3D//n3MmjULI0eOFD0XnxEPz588EZ8R3yxnxPO5rSuzSA8rbAvh/fA433Fqz0X6AI0QzdPqJKgCgQDvvfceTp06BUdHR4waNQozZ87EoEGDRNskJCQgMzMTmZmZuHz5Mt59911cvnxZm82U6ejRo7h58yZefvllvPrqqypP1iQMbmUfFydrslLpx4I9g1sMdEWTngp/Yc/Pz1d4DFUpc336TrzYISwO0QSB+sPJyQnvv/8+/vGPf7R3U9qVPuez8ObaxsYGAQEBMDc318p5RQVoAFunbhV9nfQgCbnluaLtxLdRJNgzGEkPkpBdmo2o5Ci524ufF0zsj5SQkBClC9bBnsG4HXgbKQdTsGHDBozzHCd6XNa1EqItlNHPJSYm4qOPPoJAIEBwcDDWrl0r8TxjDB999BGOHz8OMzMz7Nq1C56enu3U2ia1tbXYs2cPHj9+jLlz52Lw4MESz4cnhePGkxsITwqHVz8vvcoZ8Vz0NVVuGd6WyLqnJERX6Ws+a7UAcuXKFbi6usLZ2RkAsGjRIsTHx0vcYMfHxyMgIAAcDgdjxoxBaWkp8vPz1TKDf2sIBAKUlJTg5s2bePXVV/HKK69ovQ3SRQTxf6tSYFC0rfAX9vb6PuszWcUOVX7xIM+1VFgMDAzErl27WnXssLAwHDx4ECkpKRKPX716Fd26dWvVMQ2JPuYzAGRkZODp06ews7NDQEAAzMzMtHZuWYVoQPYqMcr0guM6cLF79m5EJUcp3F76fLZmtm3uZcd14OL0ttPAtuePib9fdJRVw4hilNHtRx+L1I2NjYiJiUFhYSEWLFgAd3f3ZtuEeoUiPCkcoV6hcLVxBaA/OSOei0VpRWo5Jn2ARlqL8ll5Wi2APHr0CH379hV97ejo2CyYZW3z6NGjdrnBfvbsGQ4ePAjGGKZMmYJx48ZpvQ3aIvyFncfjtXdT9A4VO9RHvAfS77//jpCQEInHunbtqvZz2tvbq/2Y+kjf8hkAUlNTcejQIQwaNAiBgYEa+f+hiHRRWfxrX/fnnwaq8kmmMkVtRUVxTekIvflIyyij24++FakrKyvB5/NRVFSExYsXw9XVVeZ2vu6+rc7L9iaei7w0nlqOSfeUpLUon5Wn1QIIY8376EpXq5TZBgB27NiBHTt2AADy8vLU/os7YwwlJSVgjKFbt26or6/vEMWBysrKDnGdgOFfq6WlJSoqKgA0fXIk/LcuE68id+nSpdljCQkJ+Oqrr5Ceno6ePXti/vz5WLt2LTp37gwA+PXXX/Gvf/0L9+/fh6mpKbhcLmJiYnDq1CmEh4cDeJ4nP/74I5YuXYrBgwfj7bffxocffggAsLCwwLZt23D27FmcPHkSDg4O+Pzzz7Fo0SJRO65evYqPP/4YGRkZcHNzw/r16zF//nwcO3asxVntpdXW1urE/0N15jOg+YyuqalBaWkpuFwuOnfurBNDcbTB0HNLnKFfK2U0ZbQq9KlIXV5ejpiYGNjb22Pp0qV48cUXtXp+Qjqinj17iv5tZWXV7LHffvsNYWFhSE1NRa9evbBkyRKEhoZK5PPmzZuRmZmJrl27YsiQIfjll1+QkJDQLJ+jo6OxbNmyZkNgOBwOfvrpJ5w6dQrHjx9Hjx49EBERAT8/P1E7hL3T0tLSMHDgQGzcuBFvvPEGzp49C29vb01+i0S0WgBxdHREbu7zcdF5eXno3bu3ytsAwNtvv423334bADBy5Ei1fsPq6+uxb98+PHjwADNnzkRZWZnWXpD2xuPx6FoNRHp6Orp37w4AqKioEP1bXwgr1cJ2nzhxAsHBwdi2bRteeeUVPHz4EMuXLwdjDN9++y2ePHmCoKAghIWFYcmSJaisrMSlS5dgbm6OwMBAZGVl4ffffxfdyFpaWqJr167gcDjo0qWLxPfn66+/xqZNm/DNN99g586deO+99zBlyhT069cPlZWVWLhwISZPnoy9e/fi8ePHWLlyJQDAzMxM5e+zqakphg8f3vZvWBupM58BzWb09evXwePx4OzsjLlz5+LChQsG/bMsztBzS5yhXytlNGW0KvTlQ0SBQAA+nw97e3uYmZnhwYMHePDggdqOr6sMvWArriNcq7BArS/FaWk1NTUAIGr76dOnERgYiM2bN2PcuHHIy8vDypUrUVFRgY0bN6KgoECUzzNnzkRVVRWuXr2KyspKTJs2DR988AESExNx/PhxAE2F6IqKCjDGUFdXJ/E9Cg8PR1hYGL744gvs3r0bb775JoYPH44XXngBlZWVeOONNzBx4kRs374d+fn5+OSTTwAA1dXVKn+vW1ug1moBZNSoUcjMzER2djb69OmDn3/+Gfv27ZPYZubMmfj++++xaNEiXL58GZaWllqtXNfW1mLv3r149OgR5syZgyFDhhj8Dzkh+mDjxo349NNPERQUBABwcXHB5s2b4efnh2+++QaPHz9GQ0MDfH194eTkBAASk62Zm5vDxMREohouj7+/v6ha/eWXX2Lbtm3466+/0K9fP+zduxcCgQA7d+5E165dweVy8cUXX2Dp0qXqv2gt0od8BoBLly4hMTER/fv3x8KFC2FiotW3MUKIHJTRmqUPHyLy+XzExsaivr4efn5+yMzMNOgipjhDL9iK6wjXKixQ62NxGmheoN66dSs+/fRTvPvuuwCAoUOHoqGhAX5+fti2bRsyMzPR0NCApUuXol+/fgCA0aNHi45nY2ODzp07NxvKJqtAHRAQIBrGNXjwYPz444/4+++/weVysW/fPjQ2NiI2NlbUxqqqKixdulSrBWqtLoNrYmKC77//Hq+//joGDhyIBQsWgMvlYvv27di+fTsAYNq0aXB2doarqytCQkLwww8/aK19NTU1iI2NxePHjzFv3jwMGTJEa+cmRJfpwrJs169fx8aNG2Fubi76s2TJElRVVeHJkycYOnQoXnvtNYwZMwZz587Fjz/+iKKi1k1K5uHhIfq3iYkJ7O3tUVhYCKBp0s3BgwdLjKUUf5PQV7qezwBw7tw5JCYmYuDAgVi0aBEVPwiBbuQzQBmtaeJF6vr6evz888+YOXOmxDYzZ85EbGwsGGO4dOmSVovURUVFiI6ORkNDAwIDA9GnTx+tnJcQXacLGa1MPk+cOBGDBw/uEPms9bvHadOmYdq0aRKPLV++XPRvDoeD//u//9N2s1BVVYXY2Fjw+XwsXLgQAwYM0HobCNFVurAsW2NjI0JDQzF//vxmz9nb28PY2BgnT57EmTNncP78eezcuROfffYZkpKSMHToUJXO1alTJ4mvORwOGhsbATR1MVZ1GWx9oav5zBgDj8dDUlIShgwZgtmzZ8PISKv1e0J0li7kM0AZrWniRWqBQIA333xTVKQGmrJ62rRpOH78OFxdXWFmZobo6GittO3JkyeIjY2FsbExgoKC9HZiREI0QRcyWpl8Pnr0KFJTU3Hy5EmDz2f6+AxN46NiYmJQVlaGJUuWiGbYJoQ00YVl2Tw9PZGRkSF3JnmgKWRHjx6N1157DRs2bACXy8WBAwcwdOhQdO7cGQKBoM3tGDhwIGJjY1FTUyOqYF+5cqXNxyWyMcZw+vRpnD9/HsOHD8eMGTOo+EGIGF3IZ4AyWht0sUj96NEj7N69G126dEFgYCBsbGy0en5CdJ0uZLSy+Tx27FiMHTvW4PO5wxdASktLERsbi8rKSvj5+YnGPRFCntOFZdk2bNiA6dOno1+/fliwYAFMTEyQkpKCK1eu4Ouvv8alS5dw+vRpTJgwAc7Ozvj777+Rm5srWiLQyckJDx48QHJyMl544QV0795dtIqBKpYuXYp169YhJCQEn3/+OR4/fox//etfAFpeg52ohjGGhIQEXLlyBaNGjcK0adPoe0yIFF3IZ4AyuiN6+PAh9u7dCzMzMwQGBopWniCEPKcLGa1MPh87dgwzZ85Ejx49DD6fO/THaE+fPkV0dDSqq6sREBBAxQ9CdNjrr7+OY8eO4ezZs3jppZfw0ksvYdOmTXjhhRcANM3Yff78eSxYsAD9+/fHJ598gvXr14smyps7dy6mTZuGSZMmwd7eHvv3729VO8zNzfHbb78hNTUVw4cPx6effoqwsDAATZMxEfVobGzEb7/9hitXrmDcuHFU/CBEx1FGdyzZ2dnYvXs3unfvjqCgICp+EKLDlMnnS5cuYfr06R0inztsD5Di4mLExMRAIBAgMDBQ6ysZEEIUmzdvXrMl/aZMmYIpU6bI3H7gwIFISEiQO2N3ly5dcPDgwWaP5+TkSHwtaxlB6W3GjBmDv//+W/R1fHw8OBwOXFxc5F0OUUFjYyOOHj2KW7du4ZVXXsHEiROp+EGIjqGM7rgyMzNx4MAB2Nrawt/fH+bm5u3dJEKImNbk8+HDh+WuwmJo+dxhCyCPHz8GACxbtgwODg7t3BpCiD6JiYmBs7Mz+vbti5SUFKxcuRIzZsyAnZ1dezfNIFRXVyM3NxeTJk3Cyy+/3N7NIYToGcpozbp//z7s7e3h7+8PMzOz9m4OIUSP6EI+d9gCiIeHB9zc3Fo1fokQ0rEVFBQgNDQU+fn56NmzJ9544w1s3ry5vZtlMMzNzbF8+XLKZ0JIq1BGa9aUKVPQ0NCAzp07t3dTCCF6RhfyucMWQADQzTUhpFVWr16N1atXt3czDBrlMyGktSijNYvD4VDxgxDSKrqQzx16ElRCCCGEEEIIIYR0DFQAIcSAyZqMiOgOen0I6dgoA3QbvT6EdFz086/b2vL6UAGEEAPVqVMn1NTUtHcziAINDQ0wMenQIxEJ6bAoo3VfTU0NOnXq1N7NIIRoGeWz7mtLPlMBhBAD5eDggEePHqG6upqq2DqosbERBQUFsLS0bO+mEELaAWW07mKMobq6Go8ePaKVAgnpgIT5XFdXR/msY9SRz/TRIyEGysLCAkDTks8VFRUwNTVt5xZpR21trd5ca7du3WhZRkI6KMpo3dapUyf06NFD9DoRQjoO4c99VlYWioqK2rk12qEv2Qy0PZ+pAEKIAbOwsICFhQV4PB6GDx/e3s3Rio50rYQQ/UYZTQghusnCwgLl5eXw9PRs76ZoRUfKZhoCQwghhBBCCCGEEINHBRBCCCGEEEIIIYQYPCqAEEIIIYQQQgghxOBRAYQQQgghhBBCCCEGjwoghBBCCCGEEEIIMXgcZgCLG9vZ2cHJyUljxy8qKoK9vb3Gjq9L6FoNE12rfsrJyUFxcXF7N6PNKKPVh67VMNG16idDyGhN5rMhvdYtoWs1THSt+ktRPhtEAUTTRo4ciWvXrrV3M7SCrtUw0bUSQ9aRXnO6VsNE10oMUUd6relaDRNdq2GiITCEEEIIIYQQQggxeFQAIYQQQgghhBBCiMGjAogS3n777fZugtbQtRomulZiyDrSa07XapjoWokh6kivNV2rYaJrNUw0BwghhBBCCCGEEEIMHvUAIYQQQgghhBBCiMGjAoiYxMREuLm5wdXVFZs2bWr2PGMMH374IVxdXeHh4YHk5OR2aKV6tHSte/fuhYeHBzw8PDBu3DjcvHmzHVqpHi1dq9DVq1dhbGyMgwcParF16qXMtfJ4PAwbNgxcLhdeXl5abqH6tHStZWVlmDFjBoYOHQoul4vo6Oh2aCVRF8rn5yif9RPl83OUz4aDsvk5Q8pmgPJZGuWzAWGEMcbYs2fPmLOzM7t37x6rq6tjHh4eLDU1VWKbY8eOsalTp7LGxkZ28eJF9tJLL7VTa9tGmWs9f/48e/r0KWOMsePHjxv0tQq3mzhxIvPx8WFxcXHt0NK2U+ZaS0pK2MCBA9mDBw8YY4wVFBS0R1PbTJlr3bhxI1u9ejVjjLHCwkJmbW3N6urq2qO5pI0onymfKZ/1B+Vzx0HZbJjZzBjlM+WzYecz9QD5nytXrsDV1RXOzs7o3LkzFi1ahPj4eIlt4uPjERAQAA6HgzFjxqC0tBT5+fnt1OLWU+Zax40bB2trawDAmDFjkJeX1x5NbTNlrhUA/vvf/2Lu3LlwcHBoh1aqhzLXum/fPsyZMwcvvPACAOjt9SpzrRwOBxUVFWCMobKyEjY2NjAxMWmnFpO2oHymfKZ81h+Uzx0HZbNhZjNA+Uz5bNj5TAWQ/3n06BH69u0r+trR0RGPHj1SeRt9oOp17Ny5Ez4+Ptpomtop+7oeOXIEy5cv13bz1EqZa7179y5KSkrg7e2NESNGIDY2VtvNVAtlrvX9999Heno6evfujSFDhmDbtm0wMqLI00eUz5TPlM/6g/K546BsNsxsBiifKZ8NO58Nq5zTBkzGYjgcDkflbfSBKtdx9uxZ7Ny5E+fOndN0szRCmWtduXIlNm/eDGNjY201SyOUudZnz57h+vXrOHPmDGpqajB27FiMGTMGAwYM0FYz1UKZaz1x4gSGDRuGP/74A/fu3cPkyZPx8ssvw8LCQlvNJGpC+Uz5TPmsPyifOw7KZsPMZoDymfLZsPOZCiD/4+joiNzcXNHXeXl56N27t8rb6ANlr+PWrVsIDg5GQkICbG1ttdlEtVHmWq9du4ZFixYBAIqLi3H8+HGYmJhg1qxZ2mxqmyn7f9jOzg7dunVDt27d8Morr+DmzZt6F+DKXGt0dDTWrl0LDocDV1dXvPjii8jIyMBLL72k7eaSNqJ8pnwGKJ/1BeVzx0HZbJjZDFA+Uz4beD5rdcYRHdbQ0MBefPFFdv/+fdGkMCkpKRLb/P777xITOY0aNaqdWts2ylzrgwcPmIuLCzt//nw7tVI9lLlWcYGBgXo7iZMy15qWlsZeffVV1tDQwKqqqhiXy2W3b99upxa3njLXunz5chYaGsoYY+zJkyesd+/erKioqB1aS9qK8pnymTHKZ31B+dxxUDYbZjYzRvlM+RzKGDPcfKYeIP9jYmKC77//Hq+//joEAgHefPNNcLlcbN++HQCwfPlyTJs2DcePH4erqyvMzMz0dlkgZa41IiICfD4fK1asEO1z7dq19mx2qyhzrYZCmWsdOHAgpk6dCg8PDxgZGSE4OBiDBw9u55arTplrXb9+PZYtW4YhQ4aAMYbNmzfDzs6unVtOWoPymfJZ31E+Uz4bIspmw8xmgPKZ8tmw85nDmIzBQIQQQgghhBBCCCEGxLCmdCWEEEIIIYQQQgiRgQoghBBCCCGEEEIIMXhUACGEEEIIIYQQQojBowIIIYQQQgghhBBCDB4VQAghhBBCCCGEEGLwqABCRHbt2gUOhyP60717dwwdOhTff/89nj17ptFz5+TkgMPhYNeuXaLHli1bBicnJ5WOw+PxEBYWhsbGxhaPr0saGxuxcuVK9OrVC0ZGRpg1a5bcbZ2cnMDhcLBhw4Zmz61btw4cDkfiMW9vb3A4HAQEBDTbPioqChwOBzk5OaLH6urqsHXrVgwdOhTdu3eHhYUF3N3dERgYiMzMTNF2YWFhEv9fxP8MGzZM5e8BIUQ+yuf2Q/lMCGkJZXT7oYwmqjJp7wYQ3RMXFwdHR0eUl5cjLi4OH3zwAQoLCxEREaHVdqxfvx4fffSRSvvweDyEh4dj3bp1MDJ6Xt/r1asXLl68CBcXF3U3Uy0OHjyIbdu2YcuWLRg7dixsbW1b3Gfr1q344IMPYG9vr9Q59u7di7Vr12LQoEEKt1u8eDFOnjyJ1atXY8yYMRAIBEhPT0dcXBzS0tLQv39/ie3PnTsHY2Njice6deumVJsIIaqhfNY+ymdCiLIoo7WPMpqoigogpJlhw4bB1dUVADBlyhRkZWXhu+++kxveDQ0NMDExaVY1bSt1Bm2XLl0wZswYtR1P3dLT0wEAK1eulHjTkcfLywuXLl3Cpk2bsGXLlha3HzZsGB4/foz169fj0KFDcre7f/8+jhw5gu+++07ijdPHxwcff/xxs08FAGD06NEwMaEoIUQbKJ+1j/KZEKIsymjto4wmqqIhMKRFo0aNQkVFBQoLC0Xd4H744QesXr0avXv3RpcuXVBaWgoAOHz4MMaMGQMzMzNYWVlh/vz5ePjwocTxqqursWLFCtja2sLc3BwzZ85EXl5es/PK6r5XVVWFtWvXwsXFBV26dEHPnj0xd+5cFBQUICwsDOHh4QCATp06ibqSAfK77+3ZswdDhw6Fqakp7Ozs4O/vj/z8fIltnJyc4Ofnh59//hkDBw5Et27dMHLkSJw7d06p719iYiLGjh2Lrl27wtLSErNmzcKdO3ckjh8WFgYAMDY2VqqboaOjI95991388MMPePToUYtt6NatGz7//HMcPnwY165dk7vd06dPAQA9e/aU+bwybyyEEO2hfKZ8FqJ8JkT3UEZTRgtRRusOeiVIi7Kzs2FsbAxzc3PRYxs3bsTdu3exY8cOHDlyBKampti+fTvmzp2LQYMG4eDBg/jpp5+QkpICLy8vVFRUiPZ95513EBUVhY8//hiHDx+Gm5sblixZ0mI76uvrMXnyZPznP//BsmXL8Pvvv+P777+HjY0NSkpKEBwcjLfeegtAU5eyixcv4uLFi3KPt2PHDvj7+2PgwIE4fPgwNm3ahBMnTsDLywuVlZUS2/7111/YsmULvvzySxw4cAACgQDTp08XvWnJk5iYiDfeeAPm5uY4cOAAfvzxR6SkpGDChAmi0D1y5AiWLVsGAKI2v/HGGy1+Pz7//HOYmJjgyy+/bHFbAFi+fDleeOEFrFu3Tu427u7usLCwwNq1a7Fnzx4UFBS0eFyBQIBnz55J/JFV5SaEqB/lM+WzIpTPhLQvymjKaEUoo9sJI+R/oqOjGQCWkZHBGhoa2NOnT9n27duZkZER8/X1ZYwxlp2dzQCw4cOHs8bGRtG+FRUVzMLCggUFBUkcMzs7m3Xq1Ilt3bqVMcZYRkYGMzIyYl999ZXEdsuXL2cAWHR0tOixwMBA1q9fP9HXO3fuZABYfHy83GsIDQ1lAFhDQ0Ozdogf/9mzZ8zBwYF5e3tLbPfXX38xAGzbtm2ix/r168esrKzY06dPRY9dvXqVAWB79+6V2xbGGBsxYgRzdXWVaM/9+/eZiYkJW7VqleixL774gin749ivXz+2dOlSxhhj69atY506dWJZWVlyj+Pl5cXGjx/PGGMsKiqKAWBJSUmMMcYiIyMZAJadnS3a/tdff2V2dnYMAAPAnJ2d2XvvvcfS09Mljiv8Xsv689577yl1LYQQ5VA+Uz4zRvlMiK6ijKaMZowyWl9QDxDSjLu7Ozp16gQbGxusWLECS5cuxf/7f/9PYptZs2ZJjFe8ePEiysvLsXTpUokqpqOjI9zd3fHnn38CAC5fvozGxkYsWLBA4niLFi1qsV0nT55Ez549MXPmzDZf4507d1BYWIilS5dKPD5hwgT069cPSUlJEo+PHTsW1tbWoq+HDBkCAM26JoqrqqpCcnIyFi5cKDG+78UXX8T48eObnaM1/vGPf6B79+4IDQ1Vavtly5ZhwIAB+OKLL+RuM2PGDOTk5ODw4cP44IMPYGVlhR9++AHDhw/H6dOnm21/6dIlXL16VeLP6tWrW31NhBD5KJ8pnymfCdFdlNGU0ZTRuo9mXSHNHDlyBI6OjujevTv69esHU1PTZtv06tVL4uvCwkIAwGuvvSbzmMLgE44N7NGjh8Tz0l/Lwufz0adPn5YvQAnCcXrS1wE0jd0TPi9kY2Mj8XWXLl0AALW1tXLPUVJSAsaY3HM8ePBA5XZLs7S0xOrVq/H555/js88+a3F7Y2NjREREYNGiRUhISJC7Xbdu3TB79mzMnj0bQFNAv/baa1i7dm2z8Y8jRoygCZwI0RLKZ8pnymdCdBdlNGU0ZbTuox4gpJnBgwdj5MiRcHNzkxncAJrNVi1ccmrXrl3NKplXr17Fjh07ADwPS+lxccqMk7Ozs1NqsiJlCMP4yZMnzZ578uSJUktotcTa2hocDkej5wCADz74AA4ODgrHJYpbsGABhg0bhnXr1oExptQ+Y8aMwZQpU5CWltaWphJC2ojymfJZGuUzIbqDMpoyWhpltO6hAghRi3HjxqF79+7IysrCyJEjm/1xc3MD0LTck5GREX755ReJ/X/++ecWzzFlyhQ8efIEv/32m9xthFXlmpoahcdyc3NDjx49mp33woULePDgAby8vFpsT0u6deuGESNGIC4uDgKBQPT4gwcPcOHCBbWcAwDMzMywbt06HD16FFevXm1xew6Hg3/+859ITk5utpxXRUUFysvLm+0jEAiQmZkpsxJPCNFtlM/NUT4TQnQFZXRzlNFEk6jPDVELCwsLfPPNN3jvvfdQVFQEHx8fWFpa4tGjR0hKSoK3tzeWLFkimq16w4YNaGxsxKhRo3Dq1CkcP368xXP4+fkhMjISixcvxmeffYbRo0ejoqICJ06cwMqVK+Hu7o5BgwYBALZs2QIfHx8YGxtj5MiRzY4l7Mb2zjvvwM/PD35+fnj06BG++OIL9O/fH0FBQWr5vnz55Zd44403MH36dKxYsQKVlZUIDQ2FpaUlPvnkE7WcAwBCQkLw7bff4uTJk0pt/8Ybb2D8+PE4ceKExON37tzBxIkT4e/vj0mTJsHBwQH5+fmIiopCSkoKfvjhh2bHunz5MoyNjSUeMzY2xqhRo1p/QYQQtaF8lo3ymRCiCyijZaOMJhrTnjOwEt0inME6MzNT7jbCmaAjIyNlPn/s2DHm7e3NunfvzkxNTZmLiwsLCgpiqampom2qqqrY8uXLmbW1NevWrRubMWMGO3fuXIszWDPWNFP2P/7xD/bCCy+wTp06sZ49e7K5c+eygoICxljTzNQrVqxg9vb2jMPhiGZzlp7BWmj37t3Mw8ODde7cmdnY2DA/Pz/2+PFjiW3EZ4wWB4CFhobK/V4JJSQksDFjxjBTU1NmYWHBZs6cyTIyMiS2ae0M1uJ27dolmkFanPgM1uJ4PJ5oe+EM1iUlJSw8PJy9/PLLrGfPnszExIRZWVkxb29vFhcXJ7G/ohmsu3XrptS1EEKUQ/lM+Uz5TIjuooymjKaM1h8cxpQcwEQIIYQQQgghhBCip2gOEEIIIYQQQgghhBg8KoAQQgghhBBCCCHE4FEBhBBCCCGEEEIIIQaPCiCEEEIIIYQQQggxeFQAIYQQQgghhBBCiMGjAgghhBBCCCGEEEIMHhVACCGEEEIIIYQQYvCoAEIIIYQQQgghhBCDRwUQQgghhBBCCCGEGLz/D47BDZ6ycktYAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1080x360 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def prettify(ax, title, string):\n",
    "    props = dict(facecolor='white', alpha=0.5)\n",
    "    ax.set_title(title, fontsize=16)\n",
    "    ax.set_xlabel(\"Prediction of NNSE\", fontsize=16)\n",
    "    ax.set_ylabel(\"NNSE Value\", fontsize=16)\n",
    "    lims = [\n",
    "        np.min([ax.get_xlim(), ax.get_ylim()]),  # min of both axes\n",
    "        np.max([ax.get_xlim(), ax.get_ylim()]),  # max of both axes\n",
    "    ]\n",
    "    ax.plot(lims, lims, 'k-', alpha=.5, zorder=0)\n",
    "    ax.text(0.05, 0.95, string, transform=ax.transAxes, fontsize=12, verticalalignment='top', bbox=props, zorder=0)\n",
    "    ax.legend(loc=\"lower right\", fontsize=14)\n",
    "    ax.grid()\n",
    "    return ax\n",
    "\n",
    "def getTrain(trainlists):\n",
    "    x, y = trainlists\n",
    "    x = np.array(x)\n",
    "    y = np.array(y)\n",
    "    print(x.shape)\n",
    "    print(y.shape)\n",
    "    \n",
    "    x_cfe = x[cfe_idx,:].flatten()\n",
    "    y_cfe = y[cfe_idx,:,cfe_i].flatten()\n",
    "    \n",
    "    x_lstm = x[lstm_idx].flatten()\n",
    "    y_lstm = y[lstm_idx,:, lstm_i].flatten()\n",
    "    \n",
    "    x_nwm = x[nwm_idx].flatten()\n",
    "    y_nwm = y[nwm_idx,:, nwm_i].flatten()\n",
    "    \n",
    "    return x_cfe, y_cfe, x_lstm, y_lstm, x_nwm, y_nwm\n",
    "\n",
    "pred_cfe, true_cfe, pred_lstm, true_lstm, pred_nwm, true_nwm = getTrain(trainlists)\n",
    "\n",
    "print(pred_cfe.shape)\n",
    "print(true_cfe.shape)\n",
    "\n",
    "cfestr = f\"Mean Train $R^2$: {cfe_trainr2.mean():.3f}\\nMean Test $R^2$: {cfe_testr2.mean():.3f}\\nMean Train RMSE: {np.sqrt(cfe_trainmse.mean()):.3f}\\nMean Test RMSE: {np.sqrt(cfe_testmse.mean()):.3f}\"\n",
    "lstmstr = f\"Mean Train $R^2$: {lstm_trainr2.mean():.3f}\\nMean Test $R^2$: {lstm_testr2.mean():.3f}\\nMean Train RMSE: {np.sqrt(lstm_trainmse.mean()):.3f}\\nMean Test RMSE: {np.sqrt(lstm_testmse.mean()):.3f}\"\n",
    "nwmstr = f\"Mean Train $R^2$: {nwm_trainr2.mean():.3f}\\nMean Test $R^2$: {nwm_testr2.mean():.3f}\\nMean Train RMSE: {np.sqrt(nwm_trainmse.mean()):.3f}\\nMean Test RMSE: {np.sqrt(nwm_testmse.mean()):.3f}\"\n",
    "\n",
    "fig, axs = plt.subplots(1, 3, constrained_layout=True, facecolor=\"white\")\n",
    "\n",
    "axs[0].scatter(pred_cfe, true_cfe, s=2, color=\"green\", alpha=.75, label=\"Training\")\n",
    "axs[0].scatter(testvalues[:, lstm_i], outputdataset['LSTM_nnse'], s=3, color=\"black\", label=\"Testing\")\n",
    "\n",
    "axs[1].scatter(pred_lstm, true_lstm, s=2, color=\"green\", alpha=.75, label=\"Training\")\n",
    "axs[1].scatter(testvalues[:, cfe_i], outputdataset['CFE_nnse'], s=3, color=\"black\", label=\"Testing\")\n",
    "\n",
    "axs[2].scatter(pred_nwm, true_nwm, s=2, color=\"green\", alpha=.75, label=\"Training\")\n",
    "axs[2].scatter(testvalues[:, nwm_i], outputdataset['nwm_nnse'], s=3, color=\"black\", label=\"Testing\")\n",
    "\n",
    "\n",
    "axs[0] = prettify(axs[0], \"LSTM - RFR\", lstmstr)\n",
    "axs[1] = prettify(axs[1], \"CFE - RFR\", cfestr)\n",
    "axs[2] = prettify(axs[2], \"NWM - RFR\", nwmstr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "957ef812",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'NNSE')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3gAAAEvCAYAAAAXemMoAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAA+mElEQVR4nO3de3RU9bn/8c8kQy4kXBKRahgQMeLJRUQMrQJtiVTT0FO80AoTuigNEoMSqdCCmroOto1CKrQ9SSXGlSCewkB11cRWA7JCoqVYlVbQmLQFRWAA5ZKgXEKu+/cHv0wZE+KEMOw9w/u1VlYyezbjk8njnv18rzbDMAwBAAAAAAJeiNkBAAAAAAAuDAo8AAAAAAgSFHgAAAAAECQo8AAAAAAgSFDgAQAAAECQoMADAAAAgCBhNzuAnho0aJCGDx9udhiWc/LkSUVFRZkdBgIAuYKeIF/gK3IFPUG+wFfkStc+/vhjHTlypMvnAq7AGz58uLZt22Z2GJZTXV2tiRMnmh0GAgC5gp4gX+ArcgU9Qb7AV+RK11JSUs75HEM0AQAAACBIUOABAAAAQJCgwAMAAACAIEGBBwAAAABBggIPAAAAAIIEBR4AAAAABAkKPAAAAAAIEhR4AAAAABAkKPAAAAAAIEhQ4AEAAABAkLCbHQAAAACA4GOz2cwOwYthGGaHcFFQ4AEAAAC44C5EQWWz2S6ZwuxCYYgmAAAAAAQJCjwAAAAACBIM0QQCCGPZ0RNWyhdyxdqslCsS+WJ1VsoXcgXojAIPCCCMZUdPkC/wFbmCniBfAGtjiCYAAAAABAkKPAAAAAAIEhR4AAAAABAkKPAAAAAAIEhQ4AEAAABAkKDAAwAAAIAgQYEHAAAAAEGCffAsgA1DAQAAAFwIFHgWwIahAAAAAC4EhmgCAAAAQJCgwAMAAACAIEGBBwAAAABBggIPAAAAAIIEBR4AAAAABAkKPAAAAAAIEhR4AAAAABAkKPAAAAAAIEj4tcDbsGGDrrvuOsXHx2vp0qWdnv/ss8/03e9+VzfccIOSkpK0atUqf4YDAAAAAEHNbwVeW1ubHnjgAVVUVKi2tlYul0u1tbVe5/zud79TYmKiduzYoerqai1cuFDNzc3+CgkAAAAAgprfCry3335b8fHxGjFihMLCwjR9+nSVl5d7nWOz2XT8+HEZhqETJ04oNjZWdrvdXyEBAAAAQFDzW4G3f/9+DR061PPY4XBo//79XufMmzdPdXV1iouL0/XXX6/f/va3CglhWiAAAAAAnA+/dZcZhtHpmM1m83q8ceNGjR49Wps3b9aHH36o2267TV//+tfVv39/r/OKi4tVXFwsSXK73aqurvZX2AGN9wW+IlfQE+QLfEWuoCfIF/iKXOkZvxV4DodD+/bt8zx2u92Ki4vzOmfVqlV6+OGHZbPZFB8fr6uvvlr//Oc/9dWvftXrvKysLGVlZUmSUlJSNHHiRH+FHdB4X+ArcgU9Qb7AV+QKeoJ8ga/IlZ7x23jIsWPHaufOndq9e7eam5u1bt06TZkyxeucYcOGqbKyUpL06aef6l//+pdGjBjhr5AAAAAAIKj5rQfPbrersLBQaWlpamtrU2ZmppKSklRUVCRJys7O1mOPPaZZs2bp+uuvl2EYWrZsmQYNGuSvkAAAAAAgqNmMribLWVhKSoq2bdtmdhiWY7PZupz3CHwRuYKeIF/gK3IFPUG+wFfkSte6q4lYshIAAAAAggQFHgAAAAAECXYVBwAAAOAlNjZWDQ0NZochqfNWa2aJiYlRfX292WF8KQo8AAAAAF4aGhosMfeturraMtskWKXQ/DIUeAAAAJcAK/XISNa5WQ6UXhnAVxR4AAAAlwCr9MhI9MoA/sQiKwAAAAAQJOjBAwCLYRhV1xhG1TUr5Qu5AgDmo8ADAIthGFXXrFI8WI1V8oVcAQBrYIgmAAAAAAQJCjwAAAAACBIUeAAAAAAQJCjwAAAAACBIUOABAAAAQJCgwAMAAACAIEGBBwAAAABBggIPAAAAAIIEBR4AAAAABAkKPAAAAAAIEnazAwAAAID/Gf/TX1oywOwwJEkTJana3Bg6GP/T3+wQLMkq+TJRIld6iAIPAADgEmB7/HMZhmF2GJKk6upqTZw40ewwJEk2m03GErOjsB6r5Au50nMM0QQAAACAIEGBBwAAAABBgiGaAGAxVpn3IDH3AcCFd/jUYf3mk98ouTFZgyIHmR0OEHQo8ADAYqwy70Fi7kMgsEqDwESJxgD4pOi9In3U9JGKdhTpZzf/zOxwYGE0BpwfCjwAAAKYVRoEaAyALw6fOqzyXeUyZKhsV5myb8jmxh3nRGPA+WEOHgAAAC6KoveK1G60S5LajXYV7SgyOSJY1RcbA440HjE7pIBBgQcAAAC/67hhb2lvkSS1tLdw445zojHg/FHgAQAAwO/OvmHvwI07ukJjQO9Q4AEAAMDvdhza4blh79DS3qLth7abExAsi8aA3mGRFQAAAPjdi1Ne9PxspUV5YD00BvQOBR4AAAAAy6AxoHcYogkAAHqlY68q5scAgPko8AAAXeKmHb46e68qAIC5KPAAAF3iph2+YK8qALAWCjwAQCfctMNX7FUFANZCgQcA6ISbdviCvaoAwHoo8AAAXrhph6/YqwoArIcCDwDghZt2+Iq9qgDAetgHDwDghZt2+Iq9qoDgZrPZzA7BUmJiYswOwScUeL0QGxurhoYGs8PwsMr/hDExMaqvrzc7DADniZt2AIBhGGaHIOnM/a1VYgkUFHi90NDQYJmEs9JNmFUKTQAAAOBSQ4EHAABwiaARtrNAGXYH+IoCD7hIrDSk1yof8AznBYCLxyqjjiSG3QH+RIEHXCRWGdLLcF4AAIDgxTYJAAAAABAkKPAAAAAAIEj4tcDbsGGDrrvuOsXHx2vp0qVdnlNdXa3Ro0crKSlJ3/zmN/0ZDgAAAAAENb/NwWtra9MDDzygTZs2yeFwaOzYsZoyZYoSExM95xw7dkz333+/NmzYoGHDhunQoUP+CgcAAAAAgp7fevDefvttxcfHa8SIEQoLC9P06dNVXl7udc7atWt19913a9iwYZKkwYMH+yscAAAAAAh6fuvB279/v4YOHep57HA49NZbb3md8+9//1stLS2aOHGijh8/rvnz52vmzJn+CgkAAgYrjHbGXlUAAHw5vxV4XS0H/8UbltbWVv39739XZWWlGhsbdcstt+jmm2/WyJEjvc4rLi5WcXGxJMntdqu6utpfYfeYVWI5ceKEZWKRrPO+WI0V3hdyxfqqqqrMDsEjNTXVUvGQL12jQcBbv379yJUAwN8IviJXesZvBZ7D4dC+ffs8j91ut+Li4jqdM2jQIEVFRSkqKkrf+MY3tGPHjk4FXlZWlrKysiRJKSkpltnDS5JlYrHS3maSdd4Xq7HC+0KuoKf4G1mbFfbXlNi4Gj3HtQW+Ild6xm9z8MaOHaudO3dq9+7dam5u1rp16zRlyhSvc+644w795S9/UWtrq06dOqW33npLCQkJ/goJAAAAAIKa33rw7Ha7CgsLlZaWpra2NmVmZiopKUlFRUWSpOzsbCUkJOjb3/62Ro0apZCQEN17771KTk72V0gAAAAAENRsRoCNp0hJSdG2bdvMDkOStYajWGnYnZXeFyuxyvtCrqAn+BvBV+QKeoJ8ga/Ila51VxP5daNzAAAAAMDFQ4EHAAAAAEGCAg8AAAAAggQFHgAAAAAECQo8AAAAAAgSFHgAAAAAECQo8AAAAAAgSPhto/NLgfE//aUlA8wOQ5I0UZKqzY2hg/E//c0OAQAAALgkUeD1gu3xzy2z8aLlNq9eYnYUAAAAwKWHIZoAAAAAECQo8AAAAAAgSFDgAQAAAECQoMALAodPHdZvPvmNjjQeMTsUAAAAACaiwAsCRe8V6aOmj1S0o8jsUAAAAACYiFU0A9zhU4dVvqtchgyV7SpT9g3ZGhQ5yOyw0AWrbKsxUWJLDQAAgCBFgRfgit4rUrvRLklqN9pVtKNIP7v5ZyZHha5YZVsNttQAAAAIXgzRDGAdvXct7S2SpJb2FpXtKmMuHs6J+ZoAAADBjQIvgJ3de9ehoxcP6ArzNQEAAIIbBV4A23Foh6f3rkNLe4u2H9puTkCwtC/O16QXDwAAIPgwBy+AvTjlRc/PVppXBWtiviYAAEDwowcPuAQwXxMAAODSQIEHXAKYrwkAAHBpoMADLgHM1wQAALg0MAcPuAQwXxMAAODSQA8eAAAAAASJbgu8e+65x/Pz4sWLvZ67/fbb/RMRAAAAAOC8dFvg7dy50/Pzpk2bvJ47fPiwfyICAAAAAJyXbgs8m812Xs8BAAAAAC6+bhdZOXXqlN599121t7ersbFR7777rgzDkGEYamxsvFgxAgAAAAB80G2Bd8UVV2jBggWdfu54DAAAAACwjm4LvOrq6osUBgAAAACgt7qdg/fOO+/ok08+8Tx+/vnndccdd+jBBx9UfX2934MDAAAAAPiu2wLvvvvuU1hYmCTpjTfe0MMPP6yZM2dqwIABysrKuigBAgAAAAB80+0Qzba2NsXGxkqS1q9fr6ysLE2dOlVTp07V6NGjL0Z8AAAAAAAfdduD19bWptbWVklSZWWlbr31Vs9zHccBAAAAANbQbQ+e0+nUN7/5TQ0aNEiRkZH6+te/LknatWuXBgwYcFECBAAAAAD4ptsCLzc3V5MmTdLBgwd1++23ezY3b29vV0FBwUUJEAAAAADgm26HaErSzTffrLvuuktRUVGeYyNHjtSYMWP8GhgAAACAS5PL5VJycrIkKTk5WS6Xy+SIAke3PXj9+vXz9NoZhiFJstlsam1tVXNzM/PwJM/7g/+IiYkxOwQAAACY7ELdJ3/wwQfKyMhQRkZGr16no54Jdt0WeMePH+/0+Omnn9Yzzzyju+66y6+BBQIrJYnNZrNUPOgaDQLeaAwAgEuLy+VSXl6epDO9Mrm5uXI6nSZHBX8533vToUOHqq2tTWvWrFFbW5tCQ0M1Y8YMhYaGat++fRc4yuDzpUM0JenYsWNasmSJbrjhBh0/flzvvPOOli9f7u/YgKBiGIYlvqwUS319vcl/FQDAxeJyuZSbm+tZx6GgoEC5ubkMvUMnbrdbq1evVmpqqux2u1JTU7V69Wq53W6zQwsI3RZ4R44c0SOPPKIxY8bIbrfr3Xff1S9/+UtddtllFys+AAAABIG8vDyVlJQoNTVVkpSamqqSkhJPjx6AC8NmdNN3GhUVpcsvv1w/+tGP1K9fv07PL1iwwK/BdSUlJUXbtm276P9dq2OIJnxFrqAnyBf4ily5dFhpugE5F5yGDh2q1tZWrV271jNEMyMjQ3a7nSGa/193NVG3PXg//elP9aMf/UjSmfl3X/wCAADApeV8h+UnJSVp8+bNMgxDVVVVMgxDmzdvVlJSUq+mHSD45Ofnq62tTZmZmbr99tuVmZmptrY25efnmx1aQOi2B8+K6MHrGi2n8BW5gp4gX+ArcgVfxuVyafbs2WpsbPQci4yMVElJCQutoJNRo0bp/fff9zy+/vrr9d5775kYkbV0VxN1u4rmz3/+83M+Z7PZ9Nhjj/UuMgAAAFwSnnvuOTU2NiomJkbHjh3TwIED1dDQoOeee44CD17S0tL0/vvva+7cuZo8ebJeffVVrVy5Umlpadq4caPZ4Vlet0M0o6KiOn1JUklJiZYtW3ZRAgQAAEDg27Rpk+bOnav6+npt3rxZ9fX1mjt3rjZt2mR2aLCYjlx5+umnFR0draeffppc6YFuC7yFCxd6vrKystTY2KhVq1Zp+vTp+uijj770xTds2KDrrrtO8fHxWrp06TnPe+eddxQaGqoXX3yx578BAAAALM8wDI0ZM0bJycmaNGmSkpOTNWbMGIb2ohPDMHTjjTd65cqNN95Irvio2yGaklRfX68VK1ZozZo1+uEPf6h//OMfPm1O3NbWpgceeECbNm2Sw+HQ2LFjNWXKFCUmJnY6b/HixUpLSzv/3wIAAACWt2DBApWXl3tWRrzjjjvMDgkWtWDBAr388sueXJkyZYrZIQWMbgu8n/70p/rjH/+orKwsvf/++4qOjvb5hd9++23Fx8drxIgRkqTp06ervLy8U4FXUFCgqVOn6p133jmP8AEAABAIoqKidPz4cb3wwgueeVXHjx/3TAECOkRFRenEiROaOnWqPvvsMw0YMEAnTpwgV3zU7RDN5cuX68CBA/rlL3+puLg49e/fX/3791e/fv3Uv3//bl94//79Gjp0qOexw+HQ/v37O53z0ksvKTs7uxe/AgAAAKyusbFR3/rWt1RUVKTvfve7Kioq0re+9S2vVTUBSTp16pQiIyPV0NCg9vZ2NTQ0KDIyUqdOnTI7tIDQbQ9ee3v7eb9wV2Nkv7gx5o9//GMtW7ZMoaGh3b5WcXGxiouLJUlut1vV1dXnHVcw432Br8gV9AT5Al+RK+jOsGHD1NjYKLvdrpaWFtntdjU2NmrYsGHkDrzY7XaNHz9eu3bt0t69ezVs2DDFx8fr9ddfJ1d84Ld98N58800tWbLEs5Tpk08+KUl65JFHPOdcffXVnkLwyJEj6tu3r4qLi3XnnXee83XZB69r7D8EX5Er6AnyBb4iV/Bl0tLS9NprrykmJsYz7K6hoUG33347S9/DS0jImUGGX/nKV3To0CENHjxYn376qaTedUAFk/PeB683xo4dq507d2r37t0aMmSI1q1bp7Vr13qds3v3bs/Ps2bN0n//9393W9wBAAAgMG3evFnR0dEaMGCAjh07pgEDBqilpUWbN282OzRYTGhoqEJDQ3X06FG1t7fr6NGjCgsLU1tbm9mhBYRu5+D1ht1uV2FhodLS0pSQkKB77rlHSUlJKioqUlFRkb/+swAAALCg1tZW/eEPf9Du3bu1efNm7d69W3/4wx/U2tpqdmiwmNbWVjU3Nys2NlaSFBsbq+bmZnLFR37rwZOkyZMna/LkyV7HzrWgynPPPefPUAAAAGCympoapaenez0GuhIZGanIyEiFhIR4fmaRFd/4tcADAAAApDO9MA8//LBCQ0OVmJioFStW6OGHH/b00gBnCw8PV2lpqWcfvKlTp1Lg+YgCDwAAAH5XWFio7OxsPfzww2ppaVGfPn0UHR2twsJCs0ODBbW3tyszM1N79uzRVVddxeIqPeC3OXgAAHPZbLZef13I1wFwaXM6nSoqKtLIkSMVEhKikSNHqqioSE6n0+zQYDEOh8PrM6jju8PhMDOsgEGBBwBByjCM8/6aN2+ewsPDJZ0ZJjNv3rxevR4ASGeKvJqaGlVWVqqmpobiDl3Kz89Xnz59vI716dNH+fn5JkUUWBiiCQDwkpOTo6KiIi1btkyJiYmqra3V4sWLJUkFBQUmRwcACHYdhX9eXp4kKSoqSk888QQNAj6iBw8A4OXZZ5/VtGnTVFpaqu985zsqLS3VtGnT9Oyzz5odGizI5XIpOTlZkpScnCyXy2VyRACCAb29548ePACAl6amJv31r3/1Wr0sMzNTTU1NZocGP7lQ8yQ/+OADZWRkKCMjo1evw7BeADh/9OABALzYbDZdc801ysnJUVpamnJycnTNNdewWEoQO9+5lUlJSdq8ebMMw1BVVZUMw9DmzZuVlJTEnE0AMAkFHgDAi2EYqqys1De+8Q2Vl5frG9/4hiorK7nxRid1dXWaMGGC17EJEyaorq7OpIhgdR1DeidNmsSQXsBPGKIJAPASHh6uq666SkVFRVq5cqVsNptGjhypPXv2mB0aLCYhIUGPP/64ysrKVFdXp4SEBN15551KSEgwOzRYkMvlUm5urkpKSjzDv2fPni1JzK9CJy6XS3l5eZ5rS25uLnniI3rwAABempqatGvXLg0ePFiSNHjwYO3atYs5eOgkNTVVTz75pI4cOaL29nYdOXJETz75pFJTU80ODRaUl5enkpISpaamym63KzU1VSUlJZ6VEoEOLpdL8+fP18mTJ2UYhk6ePKn58+fT4+sjCjwAgBe73a4+ffqovr5eklRfX68+ffrIbmfQB7yVlZWpX79+ioyMVEhIiCIjI9WvXz+VlZWZHRosiCG98NWiRYvU3Nws6T+LQDU3N2vRokVmhhUwKPAAAF5aW1vV3Nys2NhY2Ww2xcbGqrm5Wa2trWaHBotxu926//77FRUVJenMXlX333+/3G63yZHBihISErRlyxavY1u2bGFILzpxu92KjIxUaWmpNm7cqNLSUkVGRnJt8REFHgCgk9DQUH366acyDEOffvqpQkNDzQ4JFrVq1SoVFBRo48aNKigo0KpVq8wOCRaVm5ur2bNnq6qqSq2traqqqtLs2bOVm5trdmiwoAULFngN512wYIHZIQUMxtsAADppbW1VTEyMGhoaPN+BL7Lb7Z5hVB2am5sZzosudSyQkZOT41k4Iy8vj4Uz0KUVK1YoJSVFbW1tqqqq0ooVK8wOKWBwBQYAdCk8PFwhISEKDw83OxRYVFtbm5qbm5WWlqaWlhb16dNH4eHhamtrMzs0WNTWrVu1a9cutbe3a9euXdq6dSsFHjpxOBw6fvy4MjMztXfvXg0bNkyNjY1yOBxmhxYQGKIJAOgkKipKERERMgxDERERnjlWwNmGDBmi0NBQDRkyRDabzesx8EU5OTl6+umnNXDgQNlsNg0cOFBPP/20cnJyzA4NFpOfn6+wsDBJ8uzBGhYWpvz8fDPDChj04AEAzqlj9TLgXPr27avS0lLPvmYzZswwOyRYVFFRkQYOHCiXy+XJl+9973sqKipSQUGB2eHBQjp6dfPy8mSz2RQVFaUnnniC3l4f0YMHAPASGxurU6dO6fTp05Kk06dP69SpU4qNjTU5MljNgQMHtGzZMuXk5CgtLU05OTlatmyZDhw4YHZosKDW1lb9/ve/91o44/e//z0r9KJLTqdTNTU1qqysVE1NDcVdD1DgAQC8FBYWKjo6WkePHlV7e7uOHj2q6OhoFRYWmh0aLCYhIUEOh8PrJszhcLDsPc6ppqam28cAeo8hmgAAL2cPjamrq9PIkSOVm5tL6yk66Vj2vqSkxLPS3ezZs5WXl2d2aLCg2NhYPfLIIwoNDVViYqJWrFihRx55hNEBwAVGgQcAAM6L0+nU1q1blZ6erqamJoWHh2vOnDk0BqBLhYWFmj17thYuXOg5FhkZyegA4AJjiCYAwIvL5dL8+fN18uRJSdLJkyc1f/58uVwukyOD1bhcLq1fv15XXnmlQkJCdOWVV2r9+vXkCs4pOjpaw4cPV0hIiIYPH67o6GizQwKCDgUeAMDLokWLZLfbVVpaqo0bN6q0tFR2u12LFi0yOzRYzKJFi9TS0iLpP0uZt7S0kCvoUl5entavX6/du3ersrJSu3fv1vr16xnSC1xgFHgAAC9ut1urV6/2Wulu9erVcrvdZocGi3G73QoPD1dpaalee+01lZaWKjw8nFxBl+rq6jRhwgSvYxMmTFBdXZ1JEQHBiQIPANBJYWGhIiIilJqaqoiICObI4JxGjBih9PR03XbbbUpPT9eIESPMDgkWlZCQoC1btngd27JlC6uuoksul0vJycmaNGmSkpOTGfrdAyyyAgDwEhUVpZdffllz587V5MmT9eqrr2rlypWKiooyOzRY0JtvvtkpV4CusOoqfOVyuZSbm+vJldDQUM2ePVuSWMTJBzajY9B8gEhJSdG2bdvMDsNybDabAuxPCZOQK/gyffr0UXh4uC6//HLt2bNHV111lQ4fPqympibPfCtAOnM9iYiI0BVXXOHJlU8++USnT5/mOoMuuVwuzxYsCQkJbMGCLiUnJ6ugoECpqamqrq7WxIkTVVVVpZycHPZO/P+6q4kYogkA8NLa2qoZM2bo4MGDMgxDBw8e1IwZM9Ta2mp2aLCgjlUQbTab12OgK06nUzU1NaqsrFRNTQ3FHbrEfM3eocADAHix2+168cUXVVFRoU2bNqmiokIvvvii7HZG9cNbeHi40tLSPMN3o6KilJaWpvDwcJMjAxDImK/ZO3xaAwC89O/fXw0NDXI6nTp06JAGDx6shoYGxcTEmB0aLGbOnDl6+umndfnll8swDB05ckQul0v333+/2aEBCGDM1+wdCjwAgJeGhgZFR0ervr5ehmGovr5e0dHRamhoMDs0WMy4ceO0evVqr1yJiorSuHHjzA4NQADrGLqbk5Pjma+Zl5fHkF4fMUQTAOAlLCxMS5YsUXNzs6qqqtTc3KwlS5YoLCzM7NBgMXl5eSovL/fKlfLyclrZAfQa8zXPHz14AAAvzc3NKiws1I033ugZGlNYWKjm5mazQ4PFsBACAFgPPXgAAC+JiYnKyMhQTk6O0tLSlJOTo4yMDCUmJpodGiyGhRAA+AsbnZ8/evAAAF5yc3O73GCWYXf4IhZCAOAPbHTeOxR4AAAvTqdTW7duVXp6upqamhQeHq45c+bwoYpOyBUA/pCXl+cZSdKxyEpGRgYLrfiIAg8A4MXlcumVV15RRUWFV8vpuHHj+GCFF3IFgD/U1tbq5MmTKi0t9VxbMjMztWfPHrNDCwjMwQMAeMnLy1NJSYlSU1Nlt9uVmpqqkpISht2hE3IFgD+EhYUpJyfH69qSk5PDas4+ogcPAOCFlRHhK3IFgD+wmnPvUOABALwkJCTo8ccfV1lZmWfuw5133snKiOiEXAHgD4mJibrzzjs7zcErKyszO7SAQIEHAPCSmpqqJ598UpdffrkMw9CRI0f05JNP6v777zc7NFhMR64MHjxYknT06FFyBUCvsZpz79gMwzDMDqInUlJStG3bNrPDsBybzaYA+1PCJOQKvszQoUN19OhRtba2qqWlRX369JHdbtdll12mffv2mR0eLGTo0KGqr69XS0uLJ1f69Omj2NhYcgXdqq6u1sSJE80OAxbmcrmUl5fn6cHLzc1l8aazdFcT0YMHAPDidrt1xRVXaO3atZ6W04yMDLndbrNDg8WQKwD8xel0yul00hhwHlhFEwDQyYIFC7xWL1uwYIHZIcGiOla3S0tL86x6BwAwDz14AIBOVqxYoZSUFM/qZStWrDA7JFjUunXr9NRTTykxMVG1tbX6yU9+YnZIAHBJo8ADAHhxOBw6fvy4MjMztXfvXg0bNkyNjY1yOBxmhwaLsdvtCg8PV0FBgfbs2aOrrrpKffv2VVNTk9mhAcAly69DNDds2KDrrrtO8fHxWrp0aafn16xZo1GjRmnUqFEaN26cduzY4c9wAAA+yM/P92wm27EgT1hYmPLz880MCxbU1tamvn37eh3r27ev2traTIoIAOC3Aq+trU0PPPCAKioqVFtbK5fLpdraWq9zrr76ar3++ut677339NhjjykrK8tf4QAAfOR0OjVt2jQdPHhQhmHo4MGDmjZtGquXoZPExESNHz/eK1fGjx+vxMREs0MDgEuW3wq8t99+W/Hx8RoxYoTCwsI0ffp0lZeXe50zbtw4xcTESJJuvvlmVt0CAAtwuVx65ZVXVFFRoU2bNqmiokKvvPKKXC6X2aHBYlJTU/Xyyy9r4MCBstlsGjhwoF5++WUWWsE5uVwuJScna9KkSUpOTua6AviB3+bg7d+/X0OHDvU8djgceuutt855fklJidLT07t8rri4WMXFxZLOLMlcXV19QWMNFrwv8BW5gu48+uijmjBhgtccvAkTJujRRx/VlVdeaXZ4sJD169crMjLS61hkZKTWr1+vqVOnmhQVrKqyslIlJSX66U9/qquvvlq7d+/WwoULVVtbq0mTJpkdHiymsrJSv//97z2fQz/4wQ/IEx/5rcDraiNlm83W5blVVVUqKSnRli1bunw+KyvLM3wzJSWFvTDOgfcFviJX0J09e/ZIkkpLSz17m2VmZmrPnj3kDrwcPnxYr732mm677TbPXlWbNm3S7bffTq6gk3nz5mnNmjVKTU1VdXW1HnroIY0ePVo5OTn6xS9+YXZ4sBCXy6U1a9Z4fQ7Nnj1biYmJTBfwgd+GaDocDu3bt8/z2O12Ky4urtN57733nu69916Vl5frsssu81c4AAAfhYWFefYz69gHLycnx7PwCgCcj7q6Ok2YMMHr2IQJE1RXV2dSRLCqvLw8lZSUeH0OlZSUKC8vz+zQAoLfCryxY8dq586d2r17t5qbm7Vu3TpNmTLF65y9e/fq7rvv1v/93/9p5MiR/goFANADzc3NKiwsVFVVlVpbW1VVVaXCwkI1NzebHRosxuFw6Ic//KFXrvzwhz9kSw10KSEhodNorS1btighIcGkiGBVNAb0jt+GaNrtdhUWFiotLU1tbW3KzMxUUlKSioqKJEnZ2dn6+c9/rqNHj+r+++/3/Jtt27b5KyQAgA8SExN15513KicnR3V1dUpISFBGRobKysrMDg0Wk5+fr/nz53vN12xtbdXy5cvNDg0WlJubq9mzZ6ukpERtbW2qqqrS7Nmz6ZVBJx2NAWcv2ERjgO9sRleT5SwsJSWFIrALNputy3mPwBeRK/gyLpdLubm5npuwjrkPeXl5zH1AJy6XS3l5eZ7GgNzcXPIE50S+wBd8Dn257moiCrwgwU07fEWuwBfchKGnOhZZAXxBvuDL8DnUve5qIr8N0QRw4Z1rJVqzXodCMXg5nU45nU5uwgAACDB+W2QFwIVnGMZ5fa1du1ZXX321Nm/erE2bNmnz5s26+uqrtXbt2vN+TYq74MZmxPAVuQLgQusYollQUKCNGzeqoKBAubm5XF98RA8ecAnIy8vTZZddpkmTJskwDNlsNt10002MZUeXzjX3QRL5Ai/kCgB/OHubhI6RJCUlJcrJyeHa4gN68IBLwAcffKBt27YpOztbf/rTn5Sdna1t27bpgw8+MDs0WBD7D8FX5AoAf2CbhN6hwAMuEWPGjNEbb7yhO+64Q2+88YbGjBljdkiwqLq6Orndbq9hd263mw9WdFJXV6cXXnhBERERSk1NVUREhF544QVyBUCvJCQk6PHHH/f6HHr88cfZJsFHDNG0ACstnMG8quD1j3/8w/MzPXfoTlxcnBYvXqw1a9Z4ht3NmDFDcXFxZocGixk4cKCKi4uVn5+vxMRE1dbWatGiRRo4cKDZoQEIYKmpqVq2bJmWLVvmubYsXrxY2dnZZocWEOjBs4DzXeBi3rx5stvtWr58uSoqKrR8+XLZ7XbNmzePRTPQpZiYGK/vwLl88XrA9QFd+fzzz9W/f3/deOONstvtuvHGG9W/f399/vnnZocGIIBVVVVp8eLFKi0t1Xe+8x2VlpZq8eLFqqqqMju0wGAEmJtuusnsECwjPDzcWL58uWEYhlFVVWUYhmEsX77cCA8PNzEqWJEkQ5IRExNjhISEGDExMZ5jwBeFhIQYzz//vJGUlGSEhIQYSUlJxvPPP2+EhISYHRosRpJRWlrqlSulpaVcW3BOa9eu9cqXtWvXmh0SLCgkJMRobm42DOM/97jNzc18Dp2lu5qIIZoBrKmpSbGxsUpOTvZsAvmTn/xETU1NZocGCwoNDVVDQ4MkqaGhQaGhoWprazM5KlhRQkKCHA6HampqPKuXVVVVMfcBnYSHh6uhocErV1asWKHw8HCzQ4MFseoqfNUxB6+srMxzj3vnnXfyOeQjCrwAZrfbtXDhQr344oueC+X3vvc92e38WdFZW1ub7Ha7WltbPd+BruTm5mr27Nmem7CqqirNnj2blRHRyZw5c7R48WJJUmJiolasWME8GZwTS9/DV8zB6x0qgQDWv39/NTQ0KCMjQ4cOHdLgwYPV0NDA/CqcEwUefOF0OrV161alp6erqalJ4eHhmjNnDjdg6KSgoECS9Oijj3pyJTs723McONvZK/R29MosXryYVVfRydlz8M7OlbKyMrNDCwg2wwismfMpKSnatm2b2WFYQkhIiKKjo3X69Gm1tLSoT58+ioiI0IkTJ9Te3m52eLAQm82mmJgYzxBNSZ7HAXYJwEVwrmFUeXl5FHk4p44eGeBchg4dqtbWVq1du9ZzbcnIyJDdbte+ffvMDg8WEhoaqtOnT6tPnz6ea0tLS4siIiKYXvL/dVcTsYpmAAsLC9OoUaMUEnLmzxgSEqJRo0YpLCzM5MhgRZ999pnXiqufffaZ2SHBoti8GoC/fHFLpwu1VRSCS0JCgrZs2eJ1bMuWLczB8xFDNANYU1OT/vrXv2ru3LmaPHmyXn31Va1cudLssGBRYWFhKigo0N69ezVs2DCFhYXp9OnTZocFC6qrq9OECRO8jk2YMIFhVAB65cCBA3ruueeUk5PjGXa3bNkyzZo1y+zQYDHMBe8dCrwAZrPZNGnSJL3xxht65plnlJCQoG9961uqrKw0OzRYUFNTk/bt26f29nbPd6ArHS2nqampnmO0nALoLVboha86pgOc3RjANAHfMUQzgBmGoe3bt+vkyZMyDEMnT57U9u3bmVOFThwOhyR5xq13fO84Dpyto+W0qqpKra2tnpbT3Nxcs0MDEMC4tqAnnE6nampqVFlZqZqaGoq7HqAHL4DZ7XY1NjYqOjraM4a9sbGRbRLQyeHDh2UYhkJCQtTe3u75fvjwYbNDgwXRcgrAH7i2ABcHPXgBrH///jp9+rRycnL0yiuvKCcnR6dPn1b//v3NDg0W09TUJEmeYZkd3zuOA8D5crlcSk5O1qRJk5ScnCyXy2V2SLAwemXgK64t54+ungB27Ngx3XfffV77D2VlZemZZ54xOzQAAexc2yRI4mYMXsgVAP7gcrk0f/58RUVFSZJOnjyp+fPnS+La4gt68AJYQkKCvv/97+v06dOqqqrS6dOn9f3vf5/Jyjin6Ohor+9AV9gmAb4iVwD4w6JFi2S321VaWqqNGzeqtLRUdrtdixYtMju0gECBF8CYrIyeOnXqlNd3oCtskwBfkSsA/MHtdmv16tVejUerV6+W2+02O7SAwBDNAMZkZfTUF+fgAV1hmwT4ilwB4C9VVVV66KGHPPe4U6ZMMTukgEGBF+CcTqecTqdnPxkA6K3c3FxNmzZNUVFR2rt3r4YNG6aTJ0/qt7/9rdmhwWLIFfSUy+VSXl6e56Y9NzeXhml0Ehsbq/z8fOXn5ysxMVG1tbVatGiRYmNjzQ4tIFDgAZcQm80mwzA834FzaWpq0rFjx9Te3q79+/crMjLS7JBgUadPn/bKlYiICLNDgkWxKA981bdvX7W1tamgoEB79uzRVVddpejoaPXt29fs0AICc/CAS0RoaKinqDMMQ6GhoSZHBKtatGiRbDabhgwZopCQEA0ZMkQ2m43J7ehk0aJFio6O1saNG7Vp0yZt3LhR0dHR5Aq6xKI88NWBAweUkZGhgwcPyjAMHTx4UBkZGTpw4IDZoQUECrwAxx4h8FVERISGDx8um82m4cOH08qOc3K73YqMjPRavSwyMpLJ7ejE7XZr1qxZysnJUVpamnJycjRr1ixyBV1iUR74Ki4uTmVlZaqoqNCmTZtUUVGhsrIyxcXFmR1aQGCIZgBjqAN8FRsbq/r6ep08eVKS9PHHH3uOA11ZsGCBUlNTPfN7FyxYQK8MurRq1SqtXbvW8zmUkZFhdkiwKBblQU98cSoJU0t8Rw9eAMvLy1NGRoZXy2lGRgZDHdBJSkqKJCkkJMTre8dx4ItWrFjhtQXLihUrzA4JFmS329XU1OR1rKmpSXY77cfojO2d4KsDBw4oPz/f6x43Pz+fIZo+4gocwGpra3Xq1KlOPXgdvTNAh9dff10jR47Uzp07JZ1pBRs5cqRef/11kyODFTkcDh0/flyZmZmelREbGxvlcDjMDg0W0/HZk5mZ6VkIITQ0VG1tbWaHBgtieyf4KiEhQQ6HQzU1NZ6RJFVVVfT2+ogevAAWFhamefPmeU1WnjdvnsLCwswODRbT1NSkDz/8UE899ZQqKir01FNP6cMPP+zU8g5IUn5+vuc60jEkJiwsTPn5+WaGBQtKTEzUfffdp6ioKNlsNkVFRem+++5TYmKi2aHBopxOp2pqalRZWamamhqKO3SJ3t7eoQcvgDU3N+vJJ59UQUGBp5X9xIkTam5uNjs0WNCoUaNUWlrqaTUdNWqU3n33XbPDggV13HDl5eV5btqfeOIJbsTQSW5ubpdzwZkqAKA3nE6ntm7dqvT0dDU1NSk8PFxz5szhc8hHFHgBbMiQITp06JCOHDki6czCGWFhYRoyZIjJkcGK3n33Xc2dO1dLly7Vq6++qpUrV5odEizM6XTK6XR6hsYAXWHIHQB/cLlceuWVV1RRUeHVeDRu3DiuLz5giGYAa2hoUHNzs2JiYmSz2RQTE6Pm5mY1NDSYHRosxmazyeFwqKioSN/97ndVVFQkh8Mhm81mdmiwKLZgAQCYhT0Te4cevAB28uRJRUVFacCAAfrss880YMAANTc3e5bCBzoYhiG3262YmBg1NDRo4MCB7FOFc2ILFviKXAHgD3V1dXK73UpOTvaMDli8eDF7JvqIHrwA97Of/Uy7d+9WZWWldu/erZ/97GdmhwQLstvtCgsL04kTJyRJJ06cUFhYGEuZo0u0nMJX5AoAf4iLi9PixYtVUFCgjRs3qqCgQIsXL2ajcx9R4AW4X/3qV14rDP3qV78yOyRYUGtrq2JjY7Vx40Zt2rRJGzduVGxsrFpbW80ODRZ0dstpxxBNt9tNyyk6IVcA+AsbnZ8/mu8DWGxsrBoaGpSRkaFDhw5p8ODBamhoUGxsrNmhwYJmzZrltRDCrFmztHTpUrPDggV1tJyuWbPGM+xuxowZtJyiE3IFgD8cOHBAzz33nNd9S35+vmbNmmV2aAGBAi+AFRYWKjs7W0ePHlV7e7uOHj2qfv36qbCw0OzQYDEOh0OrV6/udBPGxtU4F1pO4StyBcCFxkbnvcMQzQDmdDo1c+ZMhYSc+TOGhIRo5syZTGxHJ/n5+WptbVVmZqbS0tKUmZmp1tZWNq5Glw4cOKC77rpL6enpuu2225Senq677rpLBw4cMDs0WMyBAweUn5+vnJwcpaWlKScnR/n5+eQKgF5ho/PeoQcvgLFHCHzVsWHos88+q/b2dh08eJANQ3FOcXFxKisr87q2MOwOXaGVHYA/sMdm79CDF8BYvQy+OrsxYNOmTaqoqNArr7zC3mY4J4bdwRe0sgPwF6fTqZqaGlVWVqqmpobirgfowQtgdXV1mjBhgtexCRMmsHoZOjm7MaCjlb2kpEQ5OTlcMNHJgQMHdN999yk9PV1NTU0KDw9XZmamnnnmGbNDg8XQyg4A1kOBF8ASEhJ0zz33qKKiwnMTlp6eztAYdMKGoeiJuLg4vfTSS15DNDMyMhiiiS45nU45nU5P4xEAwFwUeAFsyJAhKisr09y5czV58mS9+uqrWrlypW6//XazQ4PFsJQ5eqq+vl633nqr53FYWJgGDx5sYkSwKpfLpby8PE/jUW5uLj14AGAiCrwA9vrrr2vGjBl644039MwzzyghIUEzZszQiy++aHZosCDmVMFXbrdbkhQTE6Njx45p4MCBamho8BwHOrhcLuXm5qqkpMRrsS9JFHkAYBK/LrKyYcMGXXfddYqPj+9yQ2XDMPTggw8qPj5eo0aN0j/+8Q9/hhN0mpqaVFxc7DUBtbi4WE1NTWaHBothKXP01JQpU1RfX6/Nmzervr5eU6ZMMTskWBCLfQGA9fitwGtra9MDDzygiooK1dbWyuVyqba21uuciooK7dy5Uzt37lRxcbHmzp3rr3CCUnh4uIqKiryOFRUVKTw83KSIYFVnL2Xe0RjgcDiYr4lz2r59u9fKiNu3bzc7JFgQi30BgPX4rcB7++23FR8frxEjRigsLEzTp09XeXm51znl5eWaOXOmbDabbr75Zh07dkwHDx70V0hBZ86cOVq8eLFWrFih06dPa8WKFVq8eLHmzJljdmiwGJYyR0/YbDZde+21Xj2+1157rWw2m9mhwWISEhK0ZcsWr2Nbtmyh8QgATOS3OXj79+/X0KFDPY8dDofeeuutLz1n//79uvLKK/0VVlApKCiQJD366KOeVTSzs7M9x4EOLGWOnrjtttv02muvae7cuVq6dCkLOOGcOhqPOubgdTQeMUQTAMzjtwKvqwUcvtj668s5klRcXKzi4mJJZyb/V1dXX5ggg8DUqVM1depUnThxQtHR0ZLE+4MuXXnllSosLCRX8KUeeeQR1dfXa+XKlVq5cqUkKSUlRY888gg5Ay9XXnmlZsyYoczMTO3du1fDhg3TD37wA1155ZXkCrp14sQJcgQ+IVd6zm8FnsPh0L59+zyP3W53pyXZfTlHkrKyspSVlSXpzE0G++x0xv5D8BW5Al+88847ksgXfLmJEyfqF7/4BbmCHiFf4Ctypef8Ngdv7Nix2rlzp3bv3q3m5matW7eu0ypsU6ZM0fPPPy/DMPS3v/1NAwYMYHgmAAAAAJwnv/Xg2e12FRYWKi0tTW1tbcrMzFRSUpJn1cfs7GzP5tzx8fHq27evVq1a5a9wAAAAACDo+XWj88mTJ2vy5Mlex7Kzsz0/22w2/e53v/NnCAAAAABwyfDrRucAAAAAgIuHAg8AAAAAggQFHgAAAAAECQo8AAAAAAgSFHgAAAAAECRshmEYZgfRE4MGDdLw4cPNDsNyDh8+rMsvv9zsMBAAyBX0BPkCX5Er6AnyBb4iV7r28ccf68iRI10+F3AFHrqWkpKibdu2mR0GAgC5gp4gX+ArcgU9Qb7AV+RKzzFEEwAAAACCBAUeAAAAAAQJCrwgkZWVZXYICBDkCnqCfIGvyBX0BPkCX5ErPcccPAAAAAAIEvTgAQAAAECQoMALEJ988ommT5+ua665RomJiZo8ebL+/e9/KzIyUqNHj/Z8NTc367nnntPll1/udby2ttbsXwF+YrPZtHDhQs/jp556SkuWLNGxY8d02WWXqaOT/s0335TNZpPb7ZYkffbZZ4qNjVV7e7tmzZqlvn376vjx457XmT9/vmw22zmX4EVgio6O7nTsX//6lyZOnKjRo0crISFBWVlZ2rhxo+f6ER0dreuuu06jR4/WzJkzVV1dLZvNppKSEs9rvPvuu7LZbHrqqacu5q+DiyQ0NFSjR4/WDTfcoDFjxmjr1q2SzizT3fE5lJiYqOzsbLW3t+vjjz+WzWbTY4895nmNI0eOqE+fPpo3b55ZvwYugnPlSodf//rXioiI0GeffeY5xjUFDz30kH7zm994Hqelpenee+/1PF64cKFWrFjR6b73+eefNyFa66PACwCGYeiuu+7SxIkT9eGHH6q2tlZPPPGEPv30U11zzTXavn275yssLEySNG3aNK/jiYmJJv8W8Jfw8HD98Y9/7FSIDRw4UFdccYXq6uokSVu3btWNN97o+bD929/+pq997WsKCTlzGYiPj1d5ebkkqb29XVVVVRoyZMhF/E1glgcffFAPPfSQtm/frrq6OuXk5CgtLc1z/UhJSdGaNWu0fft2z4fp9ddfr/Xr13teY926dbrhhhvM+hXgZ5GRkdq+fbt27NihJ598Uo888ojnuY7Poffee0+1tbUqKyuTJI0YMUJ//vOfPee98MILSkpKutih4yLrLlckyeVyaezYsXrppZe8jnNNubSNGzfOc3/S3t6uI0eO6IMPPvA8v3XrVo0fP77Tfe/MmTPNCtnSKPACQFVVlfr06aPs7GzPsdGjR2vo0KEmRgWrsNvtysrK0q9//etOz40fP95zwdy6daseeughr8fjxo3znOt0Oj0frtXV1Ro/frzsdvtF+A1gtoMHD8rhcHgeX3/99V/6b4YNG6bTp0/r008/lWEY2rBhg9LT0/0ZJizi888/V0xMTKfjdrtd48aN065duySdudFPSEjw7F+1fv163XPPPRc1Vpjri7ny4Ycf6sSJE/rlL38pl8vldS7XlEvb2fcrH3zwgZKTk9WvXz81NDSoqalJdXV1XV530DUKvABQU1Ojm266qcvnPvzwQ0839QMPPOA5vn79eq8u7MbGxosVLkzwwAMPaM2aNV5DXiTvFrGPPvpI3//+9z03Wx2tYR2uvfZaHT58WA0NDXK5XJo+ffrF+wVgqoceeki33nqr0tPT9etf/1rHjh3z6d9973vf0wsvvKCtW7dqzJgxCg8P92+gME1jY6NGjx6t//qv/9K9997rNfSyw6lTp1RZWenVQDB9+nStW7dObrdboaGhiouLu5hhwwTd5YrL5ZLT6dTXv/51/etf/9KhQ4e8/i3XlEtXXFyc7Ha79u7dq61bt+qWW27R1772Nb355pvatm2bRo0apbCwMK/73tGjR+svf/mL2aFbEs3zAa6jq/qLpk2bpsLCwosfEEzRv39/zZw5U//7v/+ryMhIz/Hx48dr6dKl2r17t4YPH66IiAgZhqETJ07o73//u7761a96vc7dd9+tdevW6a233tIzzzxzsX8NmORHP/qR0tLStGHDBpWXl+uZZ57Rjh07vvTm6p577tG0adP0z3/+U06ns9NcGwSPjmF30pn5vDNnzlRNTY2k/zQ02mw23XHHHUpPT9fHH38sSfr2t7+txx57TF/5ylc0bdo0k6LHxXSuXLHZbFq3bp1eeuklhYSE6O6779YLL7zg1TjNNeXS1tGLt3XrVi1YsED79+/X1q1bNWDAAM+Io3Pd98IbPXgBICkpSX//+9/NDgMW9+Mf/1glJSU6efKk59i1116rhoYG/elPf9Itt9wiSbrpppu0atUqXX311Z0W3Jg+fboee+wx3XbbbZ65ebg0xMXFKTMzU+Xl5bLb7Z6b9+5cccUV6tOnjzZt2qRJkyZdhChhBbfccouOHDmiw4cPS/rPDde7776rJUuWeJ0bFhamm266ScuXL9fUqVNNiBZmOjtX3nvvPe3cuVO33Xabhg8frnXr1nUapsk15dLWMero/fffV3Jysm6++Wa9+eabnUYc4ctxBxcAbr31VjU1NenZZ5/1HHvnnXe0Z88eE6OC1cTGxuqee+7xWoVMOvMB+9vf/tZT4N1yyy36zW9+4zX/rsOwYcOUl5en+++//6LEDGvYsGGDWlpaJJ1Zsffo0aM+L7Dz85//XMuWLVNoaKg/Q4SF/POf/1RbW5suu+wyn85fuHChli1b5vP5CB5n54rL5dKSJUv08ccf6+OPP9aBAwe0f//+TvcyXFMuXePHj9ef//xnxcbGKjQ0VLGxsTp27JjefPNNzz0MfMMQzQBgs9n00ksv6cc//rGWLl2qiIgIDR8+3Gs52S9av369tmzZ4nn89NNPd3lDj+CycOHCTkNzx48fr1dffVUpKSmSzhR4H3300Tnz4b777vN7nDDPqVOnvBZUWbBggdxut+bPn6+IiAhJ0q9+9StdccUVPr0e15VLQ8e8KunMys6rV6/2+QY8KSmJ1TMvIefKlXXr1qmiosLr3Lvuukvr1q3T1772Nc8xrimXruuvv15HjhxRRkaG17ETJ05o0KBBOnHihGdIeIfMzEw9+OCDJkRrbTajY5MsAAAAAEBAY4gmAAAAAAQJCjwAAAAACBIUeAAAAAAQJCjwAAAAACBIUOABAAAAQJCgwAMAAACAIEGBBwAAAABBggIPAAAAAILE/wNUMriuIwnrhgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams['figure.figsize'] = (15, 5)\n",
    "target = outputdataset\n",
    "\n",
    "# THIS IS THE BEST PREDICTED MODEL\n",
    "# CALCULATED AS THE MAXIMUM OF THE PREDICTED PERFORMANCES\n",
    "# BEST PRED MODEL = BPM\n",
    "bpm_idx = np.argmax(testvalues, axis=1)\n",
    "bpm_perf = testvalues[np.arange(testvalues.shape[0]), bpm_idx]\n",
    "\n",
    "# THIS IS THE BEST ACTUAL MODEL\n",
    "# CALCULATED AS THE MAXIMUM OF THE ACTUAL PERFORMANCES\n",
    "# BEST ACTUAL MODEL = BAM\n",
    "target_perf = target.to_numpy()\n",
    "bam_idx = np.argmax(target_perf, axis=1)\n",
    "bam_perf = target_perf[np.arange(testvalues.shape[0]), bam_idx]\n",
    "\n",
    "\n",
    "# ACTUAL MODEL PERFORMANCE = AMP\n",
    "amp = target.to_numpy()\n",
    "\n",
    "cfe = amp[:, cfe_i]\n",
    "cfe_est = testvalues[:, cfe_i]\n",
    "\n",
    "nwm = amp[:, nwm_i]\n",
    "nwm_est = testvalues[:, nwm_i]\n",
    "\n",
    "lstm = amp[:, lstm_i]\n",
    "lstm_est = testvalues[:, lstm_i]\n",
    "\n",
    "bestpred_est = testvalues[np.arange(amp.shape[0]), bpm_idx]\n",
    "bestpred_real = amp[np.arange(amp.shape[0]), bpm_idx]\n",
    "\n",
    "bestmodel = bam_perf\n",
    "\n",
    "# perf_models = [cfe_est, cfe, lstm_est, lstm, bestpred_est, bestpred_real, bestmodel]\n",
    "# labs = [\"CFE\\nRFR\", \"CFE\\nActual\", \"LSTM\\nRFR\",\"LSTM\\nActual\", \"BPM\\nRFR\", \"BPM\\nActual\", \"BAM\"]\n",
    "\n",
    "perf_models = [cfe, nwm, lstm, bestpred_real, bestmodel, nnse]\n",
    "labs = [\"CFE\", \"NWM\", \"LSTM\", \"BPM\", \"BAM\", \"WE\"]\n",
    "\n",
    "fig, ax = plt.subplots(facecolor=\"white\")\n",
    "ax.boxplot(perf_models, labels=labs, showmeans=True)\n",
    "ax.grid()\n",
    "ax.set_ylabel(\"NNSE\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "9ea2f08d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'NNSE')"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "\n",
    "perf_models = [bestpred_real, bestmodel, nnse]\n",
    "labs = [\"BPM\", \"BAM\", \"Weighted Ensemble\"]\n",
    "# \n",
    "\n",
    "plt.boxplot(perf_models, labels=labs, showmeans=True)\n",
    "plt.ylabel(\"NNSE\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "a1949852",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hru_id</th>\n",
       "      <th>basin_id</th>\n",
       "      <th>bam</th>\n",
       "      <th>bpm</th>\n",
       "      <th>ensemble</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1022500</td>\n",
       "      <td>01022500</td>\n",
       "      <td>0.795679</td>\n",
       "      <td>0.795679</td>\n",
       "      <td>0.844059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1031500</td>\n",
       "      <td>01031500</td>\n",
       "      <td>0.866430</td>\n",
       "      <td>0.866430</td>\n",
       "      <td>0.836200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1047000</td>\n",
       "      <td>01047000</td>\n",
       "      <td>0.833751</td>\n",
       "      <td>0.833751</td>\n",
       "      <td>0.814134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1052500</td>\n",
       "      <td>01052500</td>\n",
       "      <td>0.766951</td>\n",
       "      <td>0.766951</td>\n",
       "      <td>0.756785</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1054200</td>\n",
       "      <td>01054200</td>\n",
       "      <td>0.719024</td>\n",
       "      <td>0.719024</td>\n",
       "      <td>0.690853</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    hru_id  basin_id       bam       bpm  ensemble\n",
       "0  1022500  01022500  0.795679  0.795679  0.844059\n",
       "1  1031500  01031500  0.866430  0.866430  0.836200\n",
       "2  1047000  01047000  0.833751  0.833751  0.814134\n",
       "3  1052500  01052500  0.766951  0.766951  0.756785\n",
       "4  1054200  01054200  0.719024  0.719024  0.690853"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "perf_ensemble = pd.DataFrame(basins, columns=[\"basin_id\"])\n",
    "perf_ensemble[\"ensemble\"] = nnse\n",
    "\n",
    "\n",
    "bam_df = camels_df[[\"hru_id\"]].copy()\n",
    "bam_df['basin_id'] = bam_df['hru_id'].astype(str).str.zfill(8)\n",
    "bam_df[\"bam\"] = bam_perf\n",
    "bam_df[\"bpm\"] = bestpred_real\n",
    "\n",
    "perf_df = bam_df.merge(perf_ensemble, on=\"basin_id\")\n",
    "\n",
    "perf_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "0d7569b6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "261 basins where ensemble outperformed BAM, 315 where outperformed BPM\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "perf = perf_df\n",
    "perf[\"ensemble_d\"] = perf[\"ensemble\"] - perf[\"bam\"]\n",
    "perf[\"ensemble_p\"] = perf[\"ensemble\"] - perf[\"bpm\"]\n",
    "plt.hist(perf[\"ensemble_d\"], bins=20)\n",
    "print(f'{perf[perf[\"ensemble_d\"] > 0].shape[0]} basins where ensemble outperformed BAM, {perf[perf[\"ensemble_p\"] > 0].shape[0]} where outperformed BPM')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "acab68a5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 1 2]\n",
      "[400  23  72]\n"
     ]
    }
   ],
   "source": [
    "(unique, counts) = np.unique(bpm_idx, return_counts=True)\n",
    "print(unique)\n",
    "print(counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "a30f4242",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7f557ab3b910>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ConfusionMatrixDisplay.from_predictions(bam_idx, bpm_idx, normalize='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "d6e0087f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hru_id</th>\n",
       "      <th>ann_P</th>\n",
       "      <th>lon_cen</th>\n",
       "      <th>lat_cen</th>\n",
       "      <th>AREA</th>\n",
       "      <th>elev_mean_x</th>\n",
       "      <th>ave_T</th>\n",
       "      <th>july_T</th>\n",
       "      <th>Perimeter</th>\n",
       "      <th>geometry</th>\n",
       "      <th>gauge_id_x</th>\n",
       "      <th>p_mean</th>\n",
       "      <th>pet_mean</th>\n",
       "      <th>p_seasonality</th>\n",
       "      <th>frac_snow</th>\n",
       "      <th>aridity</th>\n",
       "      <th>high_prec_freq</th>\n",
       "      <th>high_prec_dur</th>\n",
       "      <th>high_prec_timing</th>\n",
       "      <th>low_prec_freq</th>\n",
       "      <th>low_prec_dur</th>\n",
       "      <th>low_prec_timing</th>\n",
       "      <th>gauge_id_y</th>\n",
       "      <th>geol_1st_class</th>\n",
       "      <th>glim_1st_class_frac</th>\n",
       "      <th>geol_2nd_class</th>\n",
       "      <th>glim_2nd_class_frac</th>\n",
       "      <th>carbonate_rocks_frac</th>\n",
       "      <th>geol_porostiy</th>\n",
       "      <th>geol_permeability</th>\n",
       "      <th>gauge_id_x</th>\n",
       "      <th>q_mean</th>\n",
       "      <th>runoff_ratio</th>\n",
       "      <th>slope_fdc</th>\n",
       "      <th>baseflow_index</th>\n",
       "      <th>stream_elas</th>\n",
       "      <th>q5</th>\n",
       "      <th>q95</th>\n",
       "      <th>high_q_freq</th>\n",
       "      <th>high_q_dur</th>\n",
       "      <th>low_q_freq</th>\n",
       "      <th>low_q_dur</th>\n",
       "      <th>zero_q_freq</th>\n",
       "      <th>hfd_mean</th>\n",
       "      <th>gauge_id_y</th>\n",
       "      <th>huc_02</th>\n",
       "      <th>gauge_name</th>\n",
       "      <th>gauge_id_x</th>\n",
       "      <th>soil_depth_pelletier</th>\n",
       "      <th>soil_depth_statsgo</th>\n",
       "      <th>soil_porosity</th>\n",
       "      <th>soil_conductivity</th>\n",
       "      <th>max_water_content</th>\n",
       "      <th>sand_frac</th>\n",
       "      <th>silt_frac</th>\n",
       "      <th>clay_frac</th>\n",
       "      <th>water_frac</th>\n",
       "      <th>organic_frac</th>\n",
       "      <th>other_frac</th>\n",
       "      <th>gauge_id_y</th>\n",
       "      <th>gauge_lat</th>\n",
       "      <th>gauge_lon</th>\n",
       "      <th>elev_mean_y</th>\n",
       "      <th>slope_mean</th>\n",
       "      <th>area_gages2</th>\n",
       "      <th>area_geospa_fabric</th>\n",
       "      <th>gauge_id</th>\n",
       "      <th>frac_forest</th>\n",
       "      <th>lai_max</th>\n",
       "      <th>lai_diff</th>\n",
       "      <th>gvf_max</th>\n",
       "      <th>gvf_diff</th>\n",
       "      <th>dom_land_cover_frac</th>\n",
       "      <th>dom_land_cover</th>\n",
       "      <th>root_depth_50</th>\n",
       "      <th>root_depth_99</th>\n",
       "      <th>LSTM_Unnamed: 0</th>\n",
       "      <th>LSTM_basin</th>\n",
       "      <th>LSTM_NSE</th>\n",
       "      <th>LSTM_fold</th>\n",
       "      <th>LSTM_nnse</th>\n",
       "      <th>CFE_basin_id</th>\n",
       "      <th>CFE_nnse</th>\n",
       "      <th>nwm_Unnamed: 0</th>\n",
       "      <th>nwm_basin_id</th>\n",
       "      <th>nwm_nnse</th>\n",
       "      <th>weight_CFE</th>\n",
       "      <th>weight_LSTM</th>\n",
       "      <th>weight_NWM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1022500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-68.07313</td>\n",
       "      <td>44.79691</td>\n",
       "      <td>6.203873e+08</td>\n",
       "      <td>103.604200</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>312624</td>\n",
       "      <td>POLYGON ((-67.97836 44.61310, -67.97800 44.613...</td>\n",
       "      <td>1022500</td>\n",
       "      <td>3.608126</td>\n",
       "      <td>2.119256</td>\n",
       "      <td>-0.114530</td>\n",
       "      <td>0.245259</td>\n",
       "      <td>0.587356</td>\n",
       "      <td>20.55</td>\n",
       "      <td>1.205279</td>\n",
       "      <td>son</td>\n",
       "      <td>233.65</td>\n",
       "      <td>3.662226</td>\n",
       "      <td>jja</td>\n",
       "      <td>1022500</td>\n",
       "      <td>Acid plutonic rocks</td>\n",
       "      <td>0.590658</td>\n",
       "      <td>Siliciclastic sedimentary rocks</td>\n",
       "      <td>0.164618</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0710</td>\n",
       "      <td>-14.2138</td>\n",
       "      <td>1022500</td>\n",
       "      <td>2.173062</td>\n",
       "      <td>0.602269</td>\n",
       "      <td>1.776280</td>\n",
       "      <td>0.554478</td>\n",
       "      <td>1.702782</td>\n",
       "      <td>0.204734</td>\n",
       "      <td>7.123049</td>\n",
       "      <td>3.90</td>\n",
       "      <td>2.294118</td>\n",
       "      <td>65.15</td>\n",
       "      <td>17.144737</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>166.25</td>\n",
       "      <td>1022500</td>\n",
       "      <td>1</td>\n",
       "      <td>Narraguagus River at Cherryfield, Maine</td>\n",
       "      <td>1022500</td>\n",
       "      <td>17.412808</td>\n",
       "      <td>1.491846</td>\n",
       "      <td>0.415905</td>\n",
       "      <td>2.375005</td>\n",
       "      <td>0.626229</td>\n",
       "      <td>59.390156</td>\n",
       "      <td>28.080937</td>\n",
       "      <td>12.037646</td>\n",
       "      <td>1.226913</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.358472</td>\n",
       "      <td>1022500</td>\n",
       "      <td>44.60797</td>\n",
       "      <td>-67.93524</td>\n",
       "      <td>92.68</td>\n",
       "      <td>17.79072</td>\n",
       "      <td>573.60</td>\n",
       "      <td>620.38</td>\n",
       "      <td>1022500</td>\n",
       "      <td>0.9232</td>\n",
       "      <td>4.871392</td>\n",
       "      <td>3.746692</td>\n",
       "      <td>0.863936</td>\n",
       "      <td>0.337712</td>\n",
       "      <td>0.820493</td>\n",
       "      <td>Mixed Forests</td>\n",
       "      <td>0.237435</td>\n",
       "      <td>2.238444</td>\n",
       "      <td>209</td>\n",
       "      <td>1022500</td>\n",
       "      <td>0.743211</td>\n",
       "      <td>fold_4</td>\n",
       "      <td>0.795679</td>\n",
       "      <td>1022500</td>\n",
       "      <td>0.702071</td>\n",
       "      <td>0</td>\n",
       "      <td>1022500</td>\n",
       "      <td>0.778313</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.545250</td>\n",
       "      <td>0.454750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1031500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-69.58119</td>\n",
       "      <td>45.23568</td>\n",
       "      <td>7.665447e+08</td>\n",
       "      <td>304.743490</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>309614</td>\n",
       "      <td>MULTIPOLYGON (((-69.33810 45.12317, -69.33800 ...</td>\n",
       "      <td>1031500</td>\n",
       "      <td>3.522957</td>\n",
       "      <td>2.071324</td>\n",
       "      <td>0.104091</td>\n",
       "      <td>0.291836</td>\n",
       "      <td>0.587950</td>\n",
       "      <td>18.90</td>\n",
       "      <td>1.148936</td>\n",
       "      <td>son</td>\n",
       "      <td>227.35</td>\n",
       "      <td>3.473644</td>\n",
       "      <td>djf</td>\n",
       "      <td>1031500</td>\n",
       "      <td>Siliciclastic sedimentary rocks</td>\n",
       "      <td>0.448928</td>\n",
       "      <td>Metamorphics</td>\n",
       "      <td>0.443863</td>\n",
       "      <td>0.026258</td>\n",
       "      <td>0.0747</td>\n",
       "      <td>-14.8410</td>\n",
       "      <td>1031500</td>\n",
       "      <td>2.030242</td>\n",
       "      <td>0.576289</td>\n",
       "      <td>1.494019</td>\n",
       "      <td>0.445091</td>\n",
       "      <td>1.648693</td>\n",
       "      <td>0.111345</td>\n",
       "      <td>8.010503</td>\n",
       "      <td>18.90</td>\n",
       "      <td>3.286957</td>\n",
       "      <td>94.80</td>\n",
       "      <td>14.697674</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>181.00</td>\n",
       "      <td>1031500</td>\n",
       "      <td>1</td>\n",
       "      <td>Piscataquis River near Dover-Foxcroft, Maine</td>\n",
       "      <td>1031500</td>\n",
       "      <td>7.252557</td>\n",
       "      <td>1.279047</td>\n",
       "      <td>0.450236</td>\n",
       "      <td>1.373292</td>\n",
       "      <td>0.559123</td>\n",
       "      <td>35.269030</td>\n",
       "      <td>50.841232</td>\n",
       "      <td>12.654125</td>\n",
       "      <td>0.674594</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1031500</td>\n",
       "      <td>45.17501</td>\n",
       "      <td>-69.31470</td>\n",
       "      <td>247.80</td>\n",
       "      <td>29.56035</td>\n",
       "      <td>769.05</td>\n",
       "      <td>766.53</td>\n",
       "      <td>1031500</td>\n",
       "      <td>0.9548</td>\n",
       "      <td>4.903259</td>\n",
       "      <td>3.990843</td>\n",
       "      <td>0.870668</td>\n",
       "      <td>0.398619</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>Mixed Forests</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>2.400000</td>\n",
       "      <td>466</td>\n",
       "      <td>1031500</td>\n",
       "      <td>0.845839</td>\n",
       "      <td>fold_9</td>\n",
       "      <td>0.866430</td>\n",
       "      <td>1031500</td>\n",
       "      <td>0.657231</td>\n",
       "      <td>1</td>\n",
       "      <td>1031500</td>\n",
       "      <td>0.757158</td>\n",
       "      <td>0.305542</td>\n",
       "      <td>0.376230</td>\n",
       "      <td>0.318228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1047000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-70.16213</td>\n",
       "      <td>44.98744</td>\n",
       "      <td>9.049562e+08</td>\n",
       "      <td>379.779978</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>310157</td>\n",
       "      <td>POLYGON ((-70.10847 45.21669, -70.10858 45.216...</td>\n",
       "      <td>1047000</td>\n",
       "      <td>3.323146</td>\n",
       "      <td>2.090024</td>\n",
       "      <td>0.147776</td>\n",
       "      <td>0.280118</td>\n",
       "      <td>0.628929</td>\n",
       "      <td>20.10</td>\n",
       "      <td>1.165217</td>\n",
       "      <td>son</td>\n",
       "      <td>235.90</td>\n",
       "      <td>3.691706</td>\n",
       "      <td>djf</td>\n",
       "      <td>1047000</td>\n",
       "      <td>Metamorphics</td>\n",
       "      <td>0.308488</td>\n",
       "      <td>Acid plutonic rocks</td>\n",
       "      <td>0.288613</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0522</td>\n",
       "      <td>-14.4819</td>\n",
       "      <td>1047000</td>\n",
       "      <td>2.182870</td>\n",
       "      <td>0.656868</td>\n",
       "      <td>1.415939</td>\n",
       "      <td>0.473465</td>\n",
       "      <td>1.510238</td>\n",
       "      <td>0.196458</td>\n",
       "      <td>8.095148</td>\n",
       "      <td>14.95</td>\n",
       "      <td>2.577586</td>\n",
       "      <td>71.55</td>\n",
       "      <td>12.776786</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>184.80</td>\n",
       "      <td>1047000</td>\n",
       "      <td>1</td>\n",
       "      <td>Carrabassett River near North Anson, Maine</td>\n",
       "      <td>1047000</td>\n",
       "      <td>5.359655</td>\n",
       "      <td>1.392779</td>\n",
       "      <td>0.422749</td>\n",
       "      <td>2.615154</td>\n",
       "      <td>0.561181</td>\n",
       "      <td>55.163133</td>\n",
       "      <td>34.185443</td>\n",
       "      <td>10.303622</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.147867</td>\n",
       "      <td>1047000</td>\n",
       "      <td>44.86920</td>\n",
       "      <td>-69.95510</td>\n",
       "      <td>310.38</td>\n",
       "      <td>49.92122</td>\n",
       "      <td>909.10</td>\n",
       "      <td>904.94</td>\n",
       "      <td>1047000</td>\n",
       "      <td>0.9906</td>\n",
       "      <td>5.086811</td>\n",
       "      <td>4.300978</td>\n",
       "      <td>0.891383</td>\n",
       "      <td>0.445473</td>\n",
       "      <td>0.850450</td>\n",
       "      <td>Mixed Forests</td>\n",
       "      <td>0.241027</td>\n",
       "      <td>2.340180</td>\n",
       "      <td>364</td>\n",
       "      <td>1047000</td>\n",
       "      <td>0.800601</td>\n",
       "      <td>fold_7</td>\n",
       "      <td>0.833751</td>\n",
       "      <td>1047000</td>\n",
       "      <td>0.638119</td>\n",
       "      <td>2</td>\n",
       "      <td>1047000</td>\n",
       "      <td>0.525211</td>\n",
       "      <td>0.434267</td>\n",
       "      <td>0.565733</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1052500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-71.17197</td>\n",
       "      <td>44.96168</td>\n",
       "      <td>3.961103e+08</td>\n",
       "      <td>646.073604</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>172588</td>\n",
       "      <td>POLYGON ((-71.10862 45.12730, -71.10825 45.127...</td>\n",
       "      <td>1052500</td>\n",
       "      <td>3.730858</td>\n",
       "      <td>2.096423</td>\n",
       "      <td>0.152097</td>\n",
       "      <td>0.352698</td>\n",
       "      <td>0.561914</td>\n",
       "      <td>13.50</td>\n",
       "      <td>1.129707</td>\n",
       "      <td>jja</td>\n",
       "      <td>193.50</td>\n",
       "      <td>2.896707</td>\n",
       "      <td>mam</td>\n",
       "      <td>1052500</td>\n",
       "      <td>Siliciclastic sedimentary rocks</td>\n",
       "      <td>0.497458</td>\n",
       "      <td>Metamorphics</td>\n",
       "      <td>0.374062</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0711</td>\n",
       "      <td>-15.1658</td>\n",
       "      <td>1052500</td>\n",
       "      <td>2.405105</td>\n",
       "      <td>0.644652</td>\n",
       "      <td>1.301062</td>\n",
       "      <td>0.459700</td>\n",
       "      <td>1.025555</td>\n",
       "      <td>0.305965</td>\n",
       "      <td>8.669019</td>\n",
       "      <td>14.10</td>\n",
       "      <td>2.517857</td>\n",
       "      <td>58.90</td>\n",
       "      <td>7.316770</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>197.20</td>\n",
       "      <td>1052500</td>\n",
       "      <td>1</td>\n",
       "      <td>Diamond River near Wentworth Location, NH</td>\n",
       "      <td>1052500</td>\n",
       "      <td>1.301349</td>\n",
       "      <td>1.494807</td>\n",
       "      <td>0.452326</td>\n",
       "      <td>1.262995</td>\n",
       "      <td>0.615538</td>\n",
       "      <td>30.557666</td>\n",
       "      <td>52.614646</td>\n",
       "      <td>11.143326</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.675527</td>\n",
       "      <td>1052500</td>\n",
       "      <td>44.87739</td>\n",
       "      <td>-71.05749</td>\n",
       "      <td>615.70</td>\n",
       "      <td>60.05183</td>\n",
       "      <td>383.82</td>\n",
       "      <td>396.10</td>\n",
       "      <td>1052500</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>4.800830</td>\n",
       "      <td>4.124313</td>\n",
       "      <td>0.880034</td>\n",
       "      <td>0.477328</td>\n",
       "      <td>0.593588</td>\n",
       "      <td>Mixed Forests</td>\n",
       "      <td>0.225615</td>\n",
       "      <td>2.237435</td>\n",
       "      <td>415</td>\n",
       "      <td>1052500</td>\n",
       "      <td>0.696136</td>\n",
       "      <td>fold_8</td>\n",
       "      <td>0.766951</td>\n",
       "      <td>1052500</td>\n",
       "      <td>0.505420</td>\n",
       "      <td>3</td>\n",
       "      <td>1052500</td>\n",
       "      <td>0.696609</td>\n",
       "      <td>0.270973</td>\n",
       "      <td>0.378079</td>\n",
       "      <td>0.350948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1054200</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-71.05872</td>\n",
       "      <td>44.31072</td>\n",
       "      <td>1.813287e+08</td>\n",
       "      <td>629.165746</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>111995</td>\n",
       "      <td>POLYGON ((-70.97999 44.39574, -70.97963 44.395...</td>\n",
       "      <td>1054200</td>\n",
       "      <td>4.067132</td>\n",
       "      <td>2.128355</td>\n",
       "      <td>0.104960</td>\n",
       "      <td>0.299642</td>\n",
       "      <td>0.523306</td>\n",
       "      <td>17.50</td>\n",
       "      <td>1.194539</td>\n",
       "      <td>son</td>\n",
       "      <td>220.30</td>\n",
       "      <td>3.263704</td>\n",
       "      <td>mam</td>\n",
       "      <td>1054200</td>\n",
       "      <td>Metamorphics</td>\n",
       "      <td>0.871443</td>\n",
       "      <td>Siliciclastic sedimentary rocks</td>\n",
       "      <td>0.104295</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0288</td>\n",
       "      <td>-14.2147</td>\n",
       "      <td>1054200</td>\n",
       "      <td>2.731742</td>\n",
       "      <td>0.671663</td>\n",
       "      <td>1.319645</td>\n",
       "      <td>0.437050</td>\n",
       "      <td>1.357808</td>\n",
       "      <td>0.256851</td>\n",
       "      <td>10.095605</td>\n",
       "      <td>16.05</td>\n",
       "      <td>1.888235</td>\n",
       "      <td>82.20</td>\n",
       "      <td>9.963636</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>183.45</td>\n",
       "      <td>1054200</td>\n",
       "      <td>1</td>\n",
       "      <td>Wild River at Gilead, Maine</td>\n",
       "      <td>1054200</td>\n",
       "      <td>0.914754</td>\n",
       "      <td>1.477442</td>\n",
       "      <td>0.418461</td>\n",
       "      <td>2.919142</td>\n",
       "      <td>0.594526</td>\n",
       "      <td>58.376169</td>\n",
       "      <td>31.836472</td>\n",
       "      <td>9.851262</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1054200</td>\n",
       "      <td>44.39044</td>\n",
       "      <td>-70.97964</td>\n",
       "      <td>472.31</td>\n",
       "      <td>90.13951</td>\n",
       "      <td>180.98</td>\n",
       "      <td>181.33</td>\n",
       "      <td>1054200</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>5.019827</td>\n",
       "      <td>4.283551</td>\n",
       "      <td>0.886711</td>\n",
       "      <td>0.455084</td>\n",
       "      <td>0.580954</td>\n",
       "      <td>Mixed Forests</td>\n",
       "      <td>0.224857</td>\n",
       "      <td>2.232382</td>\n",
       "      <td>1</td>\n",
       "      <td>1054200</td>\n",
       "      <td>0.609227</td>\n",
       "      <td>fold_0</td>\n",
       "      <td>0.719024</td>\n",
       "      <td>1054200</td>\n",
       "      <td>0.532733</td>\n",
       "      <td>4</td>\n",
       "      <td>1054200</td>\n",
       "      <td>0.547971</td>\n",
       "      <td>0.304790</td>\n",
       "      <td>0.400186</td>\n",
       "      <td>0.295024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>490</th>\n",
       "      <td>14309500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-123.74216</td>\n",
       "      <td>42.83017</td>\n",
       "      <td>2.263143e+08</td>\n",
       "      <td>737.623894</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>117484</td>\n",
       "      <td>POLYGON ((-123.81322 42.89103, -123.81312 42.8...</td>\n",
       "      <td>14309500</td>\n",
       "      <td>4.977781</td>\n",
       "      <td>3.122204</td>\n",
       "      <td>-0.995847</td>\n",
       "      <td>0.061255</td>\n",
       "      <td>0.627228</td>\n",
       "      <td>15.10</td>\n",
       "      <td>1.776471</td>\n",
       "      <td>djf</td>\n",
       "      <td>222.65</td>\n",
       "      <td>6.893189</td>\n",
       "      <td>jja</td>\n",
       "      <td>14309500</td>\n",
       "      <td>Siliciclastic sedimentary rocks</td>\n",
       "      <td>0.863563</td>\n",
       "      <td>Basic volcanic rocks</td>\n",
       "      <td>0.095225</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.1731</td>\n",
       "      <td>-14.8976</td>\n",
       "      <td>14309500</td>\n",
       "      <td>2.457700</td>\n",
       "      <td>0.493734</td>\n",
       "      <td>1.689890</td>\n",
       "      <td>0.459455</td>\n",
       "      <td>1.220096</td>\n",
       "      <td>0.057651</td>\n",
       "      <td>10.203129</td>\n",
       "      <td>41.10</td>\n",
       "      <td>4.835294</td>\n",
       "      <td>165.55</td>\n",
       "      <td>41.911392</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>125.55</td>\n",
       "      <td>14309500</td>\n",
       "      <td>17</td>\n",
       "      <td>WEST FORK COW CREEK NEAR GLENDALE, OR</td>\n",
       "      <td>14309500</td>\n",
       "      <td>0.894595</td>\n",
       "      <td>0.894531</td>\n",
       "      <td>0.442804</td>\n",
       "      <td>1.335304</td>\n",
       "      <td>0.395703</td>\n",
       "      <td>37.751217</td>\n",
       "      <td>38.879406</td>\n",
       "      <td>23.213862</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>14309500</td>\n",
       "      <td>42.80400</td>\n",
       "      <td>-123.61091</td>\n",
       "      <td>709.83</td>\n",
       "      <td>110.42527</td>\n",
       "      <td>224.92</td>\n",
       "      <td>226.31</td>\n",
       "      <td>14309500</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>4.227902</td>\n",
       "      <td>1.986325</td>\n",
       "      <td>0.883414</td>\n",
       "      <td>0.115741</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>Evergreen Needleleaf Forest</td>\n",
       "      <td>0.170000</td>\n",
       "      <td>1.800000</td>\n",
       "      <td>52</td>\n",
       "      <td>14309500</td>\n",
       "      <td>0.797050</td>\n",
       "      <td>fold_0</td>\n",
       "      <td>0.831290</td>\n",
       "      <td>14309500</td>\n",
       "      <td>0.688977</td>\n",
       "      <td>500</td>\n",
       "      <td>14309500</td>\n",
       "      <td>0.710446</td>\n",
       "      <td>0.348384</td>\n",
       "      <td>0.348678</td>\n",
       "      <td>0.302938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>491</th>\n",
       "      <td>14316700</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-122.64677</td>\n",
       "      <td>43.45290</td>\n",
       "      <td>5.880250e+08</td>\n",
       "      <td>947.553082</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>176861</td>\n",
       "      <td>POLYGON ((-122.49936 43.47688, -122.49972 43.4...</td>\n",
       "      <td>14316700</td>\n",
       "      <td>4.543400</td>\n",
       "      <td>2.277630</td>\n",
       "      <td>-0.821172</td>\n",
       "      <td>0.176337</td>\n",
       "      <td>0.501305</td>\n",
       "      <td>14.75</td>\n",
       "      <td>1.446078</td>\n",
       "      <td>djf</td>\n",
       "      <td>214.85</td>\n",
       "      <td>6.018207</td>\n",
       "      <td>jja</td>\n",
       "      <td>14316700</td>\n",
       "      <td>Basic volcanic rocks</td>\n",
       "      <td>0.444680</td>\n",
       "      <td>Mixed sedimentary rocks</td>\n",
       "      <td>0.373506</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.1338</td>\n",
       "      <td>-13.5958</td>\n",
       "      <td>14316700</td>\n",
       "      <td>2.925937</td>\n",
       "      <td>0.643997</td>\n",
       "      <td>2.231023</td>\n",
       "      <td>0.508616</td>\n",
       "      <td>1.552484</td>\n",
       "      <td>0.141493</td>\n",
       "      <td>10.653569</td>\n",
       "      <td>16.15</td>\n",
       "      <td>2.543307</td>\n",
       "      <td>136.40</td>\n",
       "      <td>33.679012</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>135.50</td>\n",
       "      <td>14316700</td>\n",
       "      <td>17</td>\n",
       "      <td>STEAMBOAT CREEK NEAR GLIDE, OR</td>\n",
       "      <td>14316700</td>\n",
       "      <td>0.879292</td>\n",
       "      <td>1.340004</td>\n",
       "      <td>0.443107</td>\n",
       "      <td>1.288301</td>\n",
       "      <td>0.584644</td>\n",
       "      <td>37.238495</td>\n",
       "      <td>38.519396</td>\n",
       "      <td>24.363634</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>14316700</td>\n",
       "      <td>43.34984</td>\n",
       "      <td>-122.72894</td>\n",
       "      <td>952.26</td>\n",
       "      <td>119.08920</td>\n",
       "      <td>587.90</td>\n",
       "      <td>588.01</td>\n",
       "      <td>14316700</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>4.859652</td>\n",
       "      <td>2.828735</td>\n",
       "      <td>0.914354</td>\n",
       "      <td>0.171176</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>Evergreen Needleleaf Forest</td>\n",
       "      <td>0.170000</td>\n",
       "      <td>1.800000</td>\n",
       "      <td>363</td>\n",
       "      <td>14316700</td>\n",
       "      <td>0.661699</td>\n",
       "      <td>fold_6</td>\n",
       "      <td>0.747216</td>\n",
       "      <td>14316700</td>\n",
       "      <td>0.705888</td>\n",
       "      <td>501</td>\n",
       "      <td>14316700</td>\n",
       "      <td>0.766645</td>\n",
       "      <td>0.328516</td>\n",
       "      <td>0.344845</td>\n",
       "      <td>0.326639</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>492</th>\n",
       "      <td>14325000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-124.02763</td>\n",
       "      <td>42.79249</td>\n",
       "      <td>4.449257e+08</td>\n",
       "      <td>649.241611</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>166110</td>\n",
       "      <td>POLYGON ((-124.07751 42.89822, -124.07716 42.8...</td>\n",
       "      <td>14325000</td>\n",
       "      <td>6.297437</td>\n",
       "      <td>2.434652</td>\n",
       "      <td>-0.952055</td>\n",
       "      <td>0.030203</td>\n",
       "      <td>0.386610</td>\n",
       "      <td>14.60</td>\n",
       "      <td>1.467337</td>\n",
       "      <td>djf</td>\n",
       "      <td>219.05</td>\n",
       "      <td>6.240741</td>\n",
       "      <td>jja</td>\n",
       "      <td>14325000</td>\n",
       "      <td>Siliciclastic sedimentary rocks</td>\n",
       "      <td>0.903702</td>\n",
       "      <td>Basic plutonic rocks</td>\n",
       "      <td>0.062996</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.1708</td>\n",
       "      <td>-15.1799</td>\n",
       "      <td>14325000</td>\n",
       "      <td>4.072283</td>\n",
       "      <td>0.646657</td>\n",
       "      <td>2.246135</td>\n",
       "      <td>0.480769</td>\n",
       "      <td>1.127976</td>\n",
       "      <td>0.104916</td>\n",
       "      <td>15.891946</td>\n",
       "      <td>23.80</td>\n",
       "      <td>3.216216</td>\n",
       "      <td>153.35</td>\n",
       "      <td>47.921875</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>125.90</td>\n",
       "      <td>14325000</td>\n",
       "      <td>17</td>\n",
       "      <td>SOUTH FORK COQUILLE RIVER AT POWERS, OR</td>\n",
       "      <td>14325000</td>\n",
       "      <td>0.990318</td>\n",
       "      <td>0.892189</td>\n",
       "      <td>0.442249</td>\n",
       "      <td>1.425770</td>\n",
       "      <td>0.388650</td>\n",
       "      <td>38.961578</td>\n",
       "      <td>40.860260</td>\n",
       "      <td>20.068726</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>14325000</td>\n",
       "      <td>42.89150</td>\n",
       "      <td>-124.07065</td>\n",
       "      <td>656.53</td>\n",
       "      <td>124.96889</td>\n",
       "      <td>443.07</td>\n",
       "      <td>444.92</td>\n",
       "      <td>14325000</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>4.150730</td>\n",
       "      <td>1.867148</td>\n",
       "      <td>0.873517</td>\n",
       "      <td>0.115977</td>\n",
       "      <td>0.997677</td>\n",
       "      <td>Evergreen Needleleaf Forest</td>\n",
       "      <td>0.170186</td>\n",
       "      <td>1.801394</td>\n",
       "      <td>156</td>\n",
       "      <td>14325000</td>\n",
       "      <td>0.738627</td>\n",
       "      <td>fold_2</td>\n",
       "      <td>0.792787</td>\n",
       "      <td>14325000</td>\n",
       "      <td>0.674310</td>\n",
       "      <td>502</td>\n",
       "      <td>14325000</td>\n",
       "      <td>0.852417</td>\n",
       "      <td>0.322267</td>\n",
       "      <td>0.353660</td>\n",
       "      <td>0.324073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>493</th>\n",
       "      <td>14362250</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-123.12535</td>\n",
       "      <td>42.16583</td>\n",
       "      <td>4.387790e+07</td>\n",
       "      <td>939.577778</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>44972</td>\n",
       "      <td>POLYGON ((-123.15128 42.19624, -123.15118 42.1...</td>\n",
       "      <td>14362250</td>\n",
       "      <td>2.781676</td>\n",
       "      <td>3.325188</td>\n",
       "      <td>-0.985486</td>\n",
       "      <td>0.141500</td>\n",
       "      <td>1.195390</td>\n",
       "      <td>20.45</td>\n",
       "      <td>1.786026</td>\n",
       "      <td>djf</td>\n",
       "      <td>260.35</td>\n",
       "      <td>7.354520</td>\n",
       "      <td>jja</td>\n",
       "      <td>14362250</td>\n",
       "      <td>Basic volcanic rocks</td>\n",
       "      <td>0.983481</td>\n",
       "      <td>Acid plutonic rocks</td>\n",
       "      <td>0.016519</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0887</td>\n",
       "      <td>-12.5264</td>\n",
       "      <td>14362250</td>\n",
       "      <td>0.332015</td>\n",
       "      <td>0.119358</td>\n",
       "      <td>1.186040</td>\n",
       "      <td>0.518408</td>\n",
       "      <td>2.694495</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.358552</td>\n",
       "      <td>33.55</td>\n",
       "      <td>9.867647</td>\n",
       "      <td>148.05</td>\n",
       "      <td>32.538462</td>\n",
       "      <td>0.088296</td>\n",
       "      <td>138.05</td>\n",
       "      <td>14362250</td>\n",
       "      <td>17</td>\n",
       "      <td>STAR GULCH NEAR RUCH, OR.</td>\n",
       "      <td>14362250</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>0.800111</td>\n",
       "      <td>0.442872</td>\n",
       "      <td>1.363910</td>\n",
       "      <td>0.348779</td>\n",
       "      <td>37.914394</td>\n",
       "      <td>39.602460</td>\n",
       "      <td>22.404372</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>14362250</td>\n",
       "      <td>42.15401</td>\n",
       "      <td>-123.07532</td>\n",
       "      <td>875.67</td>\n",
       "      <td>109.93127</td>\n",
       "      <td>41.42</td>\n",
       "      <td>43.88</td>\n",
       "      <td>14362250</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>4.430338</td>\n",
       "      <td>2.451489</td>\n",
       "      <td>0.868294</td>\n",
       "      <td>0.117102</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>Evergreen Needleleaf Forest</td>\n",
       "      <td>0.170000</td>\n",
       "      <td>1.800000</td>\n",
       "      <td>312</td>\n",
       "      <td>14362250</td>\n",
       "      <td>-0.449090</td>\n",
       "      <td>fold_5</td>\n",
       "      <td>0.408315</td>\n",
       "      <td>14362250</td>\n",
       "      <td>0.483368</td>\n",
       "      <td>503</td>\n",
       "      <td>14362250</td>\n",
       "      <td>0.490811</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.586949</td>\n",
       "      <td>0.413051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>494</th>\n",
       "      <td>14400000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-124.00527</td>\n",
       "      <td>42.23303</td>\n",
       "      <td>7.033865e+08</td>\n",
       "      <td>668.427350</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>206584</td>\n",
       "      <td>POLYGON ((-124.02084 42.36178, -124.02073 42.3...</td>\n",
       "      <td>14400000</td>\n",
       "      <td>5.556071</td>\n",
       "      <td>2.279668</td>\n",
       "      <td>-1.015946</td>\n",
       "      <td>0.024330</td>\n",
       "      <td>0.410302</td>\n",
       "      <td>19.30</td>\n",
       "      <td>1.537849</td>\n",
       "      <td>djf</td>\n",
       "      <td>237.00</td>\n",
       "      <td>6.909621</td>\n",
       "      <td>jja</td>\n",
       "      <td>14400000</td>\n",
       "      <td>Siliciclastic sedimentary rocks</td>\n",
       "      <td>0.606239</td>\n",
       "      <td>Basic plutonic rocks</td>\n",
       "      <td>0.288975</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.1216</td>\n",
       "      <td>-14.7163</td>\n",
       "      <td>14400000</td>\n",
       "      <td>7.568103</td>\n",
       "      <td>1.362132</td>\n",
       "      <td>2.017435</td>\n",
       "      <td>0.485781</td>\n",
       "      <td>1.386228</td>\n",
       "      <td>0.250706</td>\n",
       "      <td>31.226804</td>\n",
       "      <td>26.45</td>\n",
       "      <td>3.526667</td>\n",
       "      <td>148.45</td>\n",
       "      <td>46.390625</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>123.90</td>\n",
       "      <td>14400000</td>\n",
       "      <td>17</td>\n",
       "      <td>CHETCO RIVER NEAR BROOKINGS, OR</td>\n",
       "      <td>14400000</td>\n",
       "      <td>0.929015</td>\n",
       "      <td>1.071426</td>\n",
       "      <td>0.452074</td>\n",
       "      <td>0.941310</td>\n",
       "      <td>0.494187</td>\n",
       "      <td>30.026587</td>\n",
       "      <td>42.214515</td>\n",
       "      <td>27.805122</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>14400000</td>\n",
       "      <td>42.12344</td>\n",
       "      <td>-124.18731</td>\n",
       "      <td>625.31</td>\n",
       "      <td>98.81802</td>\n",
       "      <td>702.63</td>\n",
       "      <td>703.37</td>\n",
       "      <td>14400000</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>3.218271</td>\n",
       "      <td>1.430863</td>\n",
       "      <td>0.743440</td>\n",
       "      <td>0.097405</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>Evergreen Needleleaf Forest</td>\n",
       "      <td>0.170000</td>\n",
       "      <td>1.800000</td>\n",
       "      <td>208</td>\n",
       "      <td>14400000</td>\n",
       "      <td>0.798856</td>\n",
       "      <td>fold_3</td>\n",
       "      <td>0.832540</td>\n",
       "      <td>14400000</td>\n",
       "      <td>0.836775</td>\n",
       "      <td>504</td>\n",
       "      <td>14400000</td>\n",
       "      <td>0.798946</td>\n",
       "      <td>0.329331</td>\n",
       "      <td>0.362855</td>\n",
       "      <td>0.307814</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>495 rows × 89 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       hru_id  ann_P    lon_cen   lat_cen          AREA  elev_mean_x ave_T  \\\n",
       "0     1022500    0.0  -68.07313  44.79691  6.203873e+08   103.604200  None   \n",
       "1     1031500    0.0  -69.58119  45.23568  7.665447e+08   304.743490  None   \n",
       "2     1047000    0.0  -70.16213  44.98744  9.049562e+08   379.779978  None   \n",
       "3     1052500    0.0  -71.17197  44.96168  3.961103e+08   646.073604  None   \n",
       "4     1054200    0.0  -71.05872  44.31072  1.813287e+08   629.165746  None   \n",
       "..        ...    ...        ...       ...           ...          ...   ...   \n",
       "490  14309500    0.0 -123.74216  42.83017  2.263143e+08   737.623894  None   \n",
       "491  14316700    0.0 -122.64677  43.45290  5.880250e+08   947.553082  None   \n",
       "492  14325000    0.0 -124.02763  42.79249  4.449257e+08   649.241611  None   \n",
       "493  14362250    0.0 -123.12535  42.16583  4.387790e+07   939.577778  None   \n",
       "494  14400000    0.0 -124.00527  42.23303  7.033865e+08   668.427350  None   \n",
       "\n",
       "    july_T  Perimeter                                           geometry  \\\n",
       "0     None     312624  POLYGON ((-67.97836 44.61310, -67.97800 44.613...   \n",
       "1     None     309614  MULTIPOLYGON (((-69.33810 45.12317, -69.33800 ...   \n",
       "2     None     310157  POLYGON ((-70.10847 45.21669, -70.10858 45.216...   \n",
       "3     None     172588  POLYGON ((-71.10862 45.12730, -71.10825 45.127...   \n",
       "4     None     111995  POLYGON ((-70.97999 44.39574, -70.97963 44.395...   \n",
       "..     ...        ...                                                ...   \n",
       "490   None     117484  POLYGON ((-123.81322 42.89103, -123.81312 42.8...   \n",
       "491   None     176861  POLYGON ((-122.49936 43.47688, -122.49972 43.4...   \n",
       "492   None     166110  POLYGON ((-124.07751 42.89822, -124.07716 42.8...   \n",
       "493   None      44972  POLYGON ((-123.15128 42.19624, -123.15118 42.1...   \n",
       "494   None     206584  POLYGON ((-124.02084 42.36178, -124.02073 42.3...   \n",
       "\n",
       "     gauge_id_x    p_mean  pet_mean  p_seasonality  frac_snow   aridity  \\\n",
       "0       1022500  3.608126  2.119256      -0.114530   0.245259  0.587356   \n",
       "1       1031500  3.522957  2.071324       0.104091   0.291836  0.587950   \n",
       "2       1047000  3.323146  2.090024       0.147776   0.280118  0.628929   \n",
       "3       1052500  3.730858  2.096423       0.152097   0.352698  0.561914   \n",
       "4       1054200  4.067132  2.128355       0.104960   0.299642  0.523306   \n",
       "..          ...       ...       ...            ...        ...       ...   \n",
       "490    14309500  4.977781  3.122204      -0.995847   0.061255  0.627228   \n",
       "491    14316700  4.543400  2.277630      -0.821172   0.176337  0.501305   \n",
       "492    14325000  6.297437  2.434652      -0.952055   0.030203  0.386610   \n",
       "493    14362250  2.781676  3.325188      -0.985486   0.141500  1.195390   \n",
       "494    14400000  5.556071  2.279668      -1.015946   0.024330  0.410302   \n",
       "\n",
       "     high_prec_freq  high_prec_dur high_prec_timing  low_prec_freq  \\\n",
       "0             20.55       1.205279              son         233.65   \n",
       "1             18.90       1.148936              son         227.35   \n",
       "2             20.10       1.165217              son         235.90   \n",
       "3             13.50       1.129707              jja         193.50   \n",
       "4             17.50       1.194539              son         220.30   \n",
       "..              ...            ...              ...            ...   \n",
       "490           15.10       1.776471              djf         222.65   \n",
       "491           14.75       1.446078              djf         214.85   \n",
       "492           14.60       1.467337              djf         219.05   \n",
       "493           20.45       1.786026              djf         260.35   \n",
       "494           19.30       1.537849              djf         237.00   \n",
       "\n",
       "     low_prec_dur low_prec_timing  gauge_id_y  \\\n",
       "0        3.662226             jja     1022500   \n",
       "1        3.473644             djf     1031500   \n",
       "2        3.691706             djf     1047000   \n",
       "3        2.896707             mam     1052500   \n",
       "4        3.263704             mam     1054200   \n",
       "..            ...             ...         ...   \n",
       "490      6.893189             jja    14309500   \n",
       "491      6.018207             jja    14316700   \n",
       "492      6.240741             jja    14325000   \n",
       "493      7.354520             jja    14362250   \n",
       "494      6.909621             jja    14400000   \n",
       "\n",
       "                      geol_1st_class  glim_1st_class_frac  \\\n",
       "0                Acid plutonic rocks             0.590658   \n",
       "1    Siliciclastic sedimentary rocks             0.448928   \n",
       "2                       Metamorphics             0.308488   \n",
       "3    Siliciclastic sedimentary rocks             0.497458   \n",
       "4                       Metamorphics             0.871443   \n",
       "..                               ...                  ...   \n",
       "490  Siliciclastic sedimentary rocks             0.863563   \n",
       "491             Basic volcanic rocks             0.444680   \n",
       "492  Siliciclastic sedimentary rocks             0.903702   \n",
       "493             Basic volcanic rocks             0.983481   \n",
       "494  Siliciclastic sedimentary rocks             0.606239   \n",
       "\n",
       "                      geol_2nd_class  glim_2nd_class_frac  \\\n",
       "0    Siliciclastic sedimentary rocks             0.164618   \n",
       "1                       Metamorphics             0.443863   \n",
       "2                Acid plutonic rocks             0.288613   \n",
       "3                       Metamorphics             0.374062   \n",
       "4    Siliciclastic sedimentary rocks             0.104295   \n",
       "..                               ...                  ...   \n",
       "490             Basic volcanic rocks             0.095225   \n",
       "491          Mixed sedimentary rocks             0.373506   \n",
       "492             Basic plutonic rocks             0.062996   \n",
       "493              Acid plutonic rocks             0.016519   \n",
       "494             Basic plutonic rocks             0.288975   \n",
       "\n",
       "     carbonate_rocks_frac  geol_porostiy  geol_permeability  gauge_id_x  \\\n",
       "0                0.000000         0.0710           -14.2138     1022500   \n",
       "1                0.026258         0.0747           -14.8410     1031500   \n",
       "2                0.000000         0.0522           -14.4819     1047000   \n",
       "3                0.000000         0.0711           -15.1658     1052500   \n",
       "4                0.000000         0.0288           -14.2147     1054200   \n",
       "..                    ...            ...                ...         ...   \n",
       "490              0.000000         0.1731           -14.8976    14309500   \n",
       "491              0.000000         0.1338           -13.5958    14316700   \n",
       "492              0.000000         0.1708           -15.1799    14325000   \n",
       "493              0.000000         0.0887           -12.5264    14362250   \n",
       "494              0.000000         0.1216           -14.7163    14400000   \n",
       "\n",
       "       q_mean  runoff_ratio  slope_fdc  baseflow_index  stream_elas        q5  \\\n",
       "0    2.173062      0.602269   1.776280        0.554478     1.702782  0.204734   \n",
       "1    2.030242      0.576289   1.494019        0.445091     1.648693  0.111345   \n",
       "2    2.182870      0.656868   1.415939        0.473465     1.510238  0.196458   \n",
       "3    2.405105      0.644652   1.301062        0.459700     1.025555  0.305965   \n",
       "4    2.731742      0.671663   1.319645        0.437050     1.357808  0.256851   \n",
       "..        ...           ...        ...             ...          ...       ...   \n",
       "490  2.457700      0.493734   1.689890        0.459455     1.220096  0.057651   \n",
       "491  2.925937      0.643997   2.231023        0.508616     1.552484  0.141493   \n",
       "492  4.072283      0.646657   2.246135        0.480769     1.127976  0.104916   \n",
       "493  0.332015      0.119358   1.186040        0.518408     2.694495  0.000000   \n",
       "494  7.568103      1.362132   2.017435        0.485781     1.386228  0.250706   \n",
       "\n",
       "           q95  high_q_freq  high_q_dur  low_q_freq  low_q_dur  zero_q_freq  \\\n",
       "0     7.123049         3.90    2.294118       65.15  17.144737     0.000000   \n",
       "1     8.010503        18.90    3.286957       94.80  14.697674     0.000000   \n",
       "2     8.095148        14.95    2.577586       71.55  12.776786     0.000000   \n",
       "3     8.669019        14.10    2.517857       58.90   7.316770     0.000000   \n",
       "4    10.095605        16.05    1.888235       82.20   9.963636     0.000000   \n",
       "..         ...          ...         ...         ...        ...          ...   \n",
       "490  10.203129        41.10    4.835294      165.55  41.911392     0.000000   \n",
       "491  10.653569        16.15    2.543307      136.40  33.679012     0.000000   \n",
       "492  15.891946        23.80    3.216216      153.35  47.921875     0.000000   \n",
       "493   1.358552        33.55    9.867647      148.05  32.538462     0.088296   \n",
       "494  31.226804        26.45    3.526667      148.45  46.390625     0.000000   \n",
       "\n",
       "     hfd_mean  gauge_id_y  huc_02  \\\n",
       "0      166.25     1022500       1   \n",
       "1      181.00     1031500       1   \n",
       "2      184.80     1047000       1   \n",
       "3      197.20     1052500       1   \n",
       "4      183.45     1054200       1   \n",
       "..        ...         ...     ...   \n",
       "490    125.55    14309500      17   \n",
       "491    135.50    14316700      17   \n",
       "492    125.90    14325000      17   \n",
       "493    138.05    14362250      17   \n",
       "494    123.90    14400000      17   \n",
       "\n",
       "                                       gauge_name  gauge_id_x  \\\n",
       "0         Narraguagus River at Cherryfield, Maine     1022500   \n",
       "1    Piscataquis River near Dover-Foxcroft, Maine     1031500   \n",
       "2      Carrabassett River near North Anson, Maine     1047000   \n",
       "3       Diamond River near Wentworth Location, NH     1052500   \n",
       "4                     Wild River at Gilead, Maine     1054200   \n",
       "..                                            ...         ...   \n",
       "490         WEST FORK COW CREEK NEAR GLENDALE, OR    14309500   \n",
       "491                STEAMBOAT CREEK NEAR GLIDE, OR    14316700   \n",
       "492       SOUTH FORK COQUILLE RIVER AT POWERS, OR    14325000   \n",
       "493                     STAR GULCH NEAR RUCH, OR.    14362250   \n",
       "494               CHETCO RIVER NEAR BROOKINGS, OR    14400000   \n",
       "\n",
       "     soil_depth_pelletier  soil_depth_statsgo  soil_porosity  \\\n",
       "0               17.412808            1.491846       0.415905   \n",
       "1                7.252557            1.279047       0.450236   \n",
       "2                5.359655            1.392779       0.422749   \n",
       "3                1.301349            1.494807       0.452326   \n",
       "4                0.914754            1.477442       0.418461   \n",
       "..                    ...                 ...            ...   \n",
       "490              0.894595            0.894531       0.442804   \n",
       "491              0.879292            1.340004       0.443107   \n",
       "492              0.990318            0.892189       0.442249   \n",
       "493              0.625000            0.800111       0.442872   \n",
       "494              0.929015            1.071426       0.452074   \n",
       "\n",
       "     soil_conductivity  max_water_content  sand_frac  silt_frac  clay_frac  \\\n",
       "0             2.375005           0.626229  59.390156  28.080937  12.037646   \n",
       "1             1.373292           0.559123  35.269030  50.841232  12.654125   \n",
       "2             2.615154           0.561181  55.163133  34.185443  10.303622   \n",
       "3             1.262995           0.615538  30.557666  52.614646  11.143326   \n",
       "4             2.919142           0.594526  58.376169  31.836472   9.851262   \n",
       "..                 ...                ...        ...        ...        ...   \n",
       "490           1.335304           0.395703  37.751217  38.879406  23.213862   \n",
       "491           1.288301           0.584644  37.238495  38.519396  24.363634   \n",
       "492           1.425770           0.388650  38.961578  40.860260  20.068726   \n",
       "493           1.363910           0.348779  37.914394  39.602460  22.404372   \n",
       "494           0.941310           0.494187  30.026587  42.214515  27.805122   \n",
       "\n",
       "     water_frac  organic_frac  other_frac  gauge_id_y  gauge_lat  gauge_lon  \\\n",
       "0      1.226913           0.0    0.358472     1022500   44.60797  -67.93524   \n",
       "1      0.674594           0.0    0.000000     1031500   45.17501  -69.31470   \n",
       "2      0.000000           0.0    0.147867     1047000   44.86920  -69.95510   \n",
       "3      0.000000           0.0    5.675527     1052500   44.87739  -71.05749   \n",
       "4      0.000000           0.0    0.000000     1054200   44.39044  -70.97964   \n",
       "..          ...           ...         ...         ...        ...        ...   \n",
       "490    0.000000           0.0    0.000000    14309500   42.80400 -123.61091   \n",
       "491    0.000000           0.0    0.000000    14316700   43.34984 -122.72894   \n",
       "492    0.000000           0.0    0.000000    14325000   42.89150 -124.07065   \n",
       "493    0.000000           0.0    0.000000    14362250   42.15401 -123.07532   \n",
       "494    0.000000           0.0    0.000000    14400000   42.12344 -124.18731   \n",
       "\n",
       "     elev_mean_y  slope_mean  area_gages2  area_geospa_fabric  gauge_id  \\\n",
       "0          92.68    17.79072       573.60              620.38   1022500   \n",
       "1         247.80    29.56035       769.05              766.53   1031500   \n",
       "2         310.38    49.92122       909.10              904.94   1047000   \n",
       "3         615.70    60.05183       383.82              396.10   1052500   \n",
       "4         472.31    90.13951       180.98              181.33   1054200   \n",
       "..           ...         ...          ...                 ...       ...   \n",
       "490       709.83   110.42527       224.92              226.31  14309500   \n",
       "491       952.26   119.08920       587.90              588.01  14316700   \n",
       "492       656.53   124.96889       443.07              444.92  14325000   \n",
       "493       875.67   109.93127        41.42               43.88  14362250   \n",
       "494       625.31    98.81802       702.63              703.37  14400000   \n",
       "\n",
       "     frac_forest   lai_max  lai_diff   gvf_max  gvf_diff  dom_land_cover_frac  \\\n",
       "0         0.9232  4.871392  3.746692  0.863936  0.337712             0.820493   \n",
       "1         0.9548  4.903259  3.990843  0.870668  0.398619             1.000000   \n",
       "2         0.9906  5.086811  4.300978  0.891383  0.445473             0.850450   \n",
       "3         1.0000  4.800830  4.124313  0.880034  0.477328             0.593588   \n",
       "4         1.0000  5.019827  4.283551  0.886711  0.455084             0.580954   \n",
       "..           ...       ...       ...       ...       ...                  ...   \n",
       "490       1.0000  4.227902  1.986325  0.883414  0.115741             1.000000   \n",
       "491       1.0000  4.859652  2.828735  0.914354  0.171176             1.000000   \n",
       "492       1.0000  4.150730  1.867148  0.873517  0.115977             0.997677   \n",
       "493       1.0000  4.430338  2.451489  0.868294  0.117102             1.000000   \n",
       "494       1.0000  3.218271  1.430863  0.743440  0.097405             1.000000   \n",
       "\n",
       "                      dom_land_cover  root_depth_50  root_depth_99  \\\n",
       "0                      Mixed Forests       0.237435       2.238444   \n",
       "1                      Mixed Forests       0.250000       2.400000   \n",
       "2                      Mixed Forests       0.241027       2.340180   \n",
       "3                      Mixed Forests       0.225615       2.237435   \n",
       "4                      Mixed Forests       0.224857       2.232382   \n",
       "..                               ...            ...            ...   \n",
       "490      Evergreen Needleleaf Forest       0.170000       1.800000   \n",
       "491      Evergreen Needleleaf Forest       0.170000       1.800000   \n",
       "492      Evergreen Needleleaf Forest       0.170186       1.801394   \n",
       "493      Evergreen Needleleaf Forest       0.170000       1.800000   \n",
       "494      Evergreen Needleleaf Forest       0.170000       1.800000   \n",
       "\n",
       "     LSTM_Unnamed: 0  LSTM_basin  LSTM_NSE LSTM_fold  LSTM_nnse  CFE_basin_id  \\\n",
       "0                209     1022500  0.743211    fold_4   0.795679       1022500   \n",
       "1                466     1031500  0.845839    fold_9   0.866430       1031500   \n",
       "2                364     1047000  0.800601    fold_7   0.833751       1047000   \n",
       "3                415     1052500  0.696136    fold_8   0.766951       1052500   \n",
       "4                  1     1054200  0.609227    fold_0   0.719024       1054200   \n",
       "..               ...         ...       ...       ...        ...           ...   \n",
       "490               52    14309500  0.797050    fold_0   0.831290      14309500   \n",
       "491              363    14316700  0.661699    fold_6   0.747216      14316700   \n",
       "492              156    14325000  0.738627    fold_2   0.792787      14325000   \n",
       "493              312    14362250 -0.449090    fold_5   0.408315      14362250   \n",
       "494              208    14400000  0.798856    fold_3   0.832540      14400000   \n",
       "\n",
       "     CFE_nnse  nwm_Unnamed: 0  nwm_basin_id  nwm_nnse  weight_CFE  \\\n",
       "0    0.702071               0       1022500  0.778313    0.000000   \n",
       "1    0.657231               1       1031500  0.757158    0.305542   \n",
       "2    0.638119               2       1047000  0.525211    0.434267   \n",
       "3    0.505420               3       1052500  0.696609    0.270973   \n",
       "4    0.532733               4       1054200  0.547971    0.304790   \n",
       "..        ...             ...           ...       ...         ...   \n",
       "490  0.688977             500      14309500  0.710446    0.348384   \n",
       "491  0.705888             501      14316700  0.766645    0.328516   \n",
       "492  0.674310             502      14325000  0.852417    0.322267   \n",
       "493  0.483368             503      14362250  0.490811    0.000000   \n",
       "494  0.836775             504      14400000  0.798946    0.329331   \n",
       "\n",
       "     weight_LSTM  weight_NWM  \n",
       "0       0.545250    0.454750  \n",
       "1       0.376230    0.318228  \n",
       "2       0.565733    0.000000  \n",
       "3       0.378079    0.350948  \n",
       "4       0.400186    0.295024  \n",
       "..           ...         ...  \n",
       "490     0.348678    0.302938  \n",
       "491     0.344845    0.326639  \n",
       "492     0.353660    0.324073  \n",
       "493     0.586949    0.413051  \n",
       "494     0.362855    0.307814  \n",
       "\n",
       "[495 rows x 89 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "camels_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "5db75e25",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "261"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.count_nonzero(perf_df[\"ensemble\"].to_numpy() > perf_df[\"bam\"].to_numpy())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "735caeab",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f557a48a190>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams['figure.figsize'] = (10, 10)\n",
    "\n",
    "ensemblelist = list()\n",
    "lstmlist = list()\n",
    "cfelist = list()\n",
    "nwmlist = list()\n",
    "bpmlist = list()\n",
    "bamlist = list()\n",
    "quantiles = np.arange(0, 1, 0.02)\n",
    "for i in quantiles:\n",
    "    ensemblelist.append(perf_df.ensemble.quantile(i))\n",
    "    bpmlist.append(perf_df.bpm.quantile(i))\n",
    "    lstmlist.append(camels_df.LSTM_nnse.quantile(i))\n",
    "    cfelist.append(camels_df.CFE_nnse.quantile(i))\n",
    "    nwmlist.append(camels_df.nwm_nnse.quantile(i))\n",
    "    bamlist.append(perf_df.bam.quantile(i))\n",
    "\n",
    "    \n",
    "quantiles = quantiles*100\n",
    "plt.plot(quantiles, ensemblelist, \"Orange\",label=\"Weighted Ensemble\")\n",
    "plt.plot(quantiles, lstmlist, \"Blue\", label=\"Long-Short Term Memory\")\n",
    "plt.plot(quantiles, cfelist, \"Red\", label=\"Conceptual Functional Equivalent\")\n",
    "plt.plot(quantiles, nwmlist, \"Green\", label=\"National Water Model\")\n",
    "plt.plot(quantiles, bpmlist, \"Purple\", label=\"Best Predicted Model\")\n",
    "plt.plot(quantiles, bamlist, \"brown\", linestyle=\"--\", label=\"Best Actual Model\")\n",
    "plt.xlabel(\"Percentile\")\n",
    "plt.ylabel(\"NNSE\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "6df9d342",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f557afc9e80>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmwAAAJVCAYAAABj4KlAAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOzdd3zM9x/A8ddF9hAiS4YEEcSOVXuvolbtkti0Ommt/kppVVF0KFVitJSiKCV2IlQRe8QWMxI7iMz7/P44OU6CxMhlvJ+PRx539/187vt93/eIt8/UKKUUQgghhBAi2zIxdgBCCCGEEOLZJGETQgghhMjmJGETQgghhMjmJGETQgghhMjmJGETQgghhMjmJGETQgghhMjmTI0dwOvk6OiIt7e3scMQQgghhHiuyMhIrl+/nm5Zrk7YvL29CQ8PN3YYQgghhBDPVaVKlaeWSZeoEEIIIUQ2JwmbEEIIIUQ2JwmbEEIIIUQ2JwmbEEIIIUQ2JwmbEEIIIUQ2l6tniT5PbGwsMTExJCUlGTsUIUQuZmZmhrOzM/nz5zd2KEKIHCrPJmyxsbFER0fj7u6OlZUVGo3G2CEJIXIhpRQPHjzg8uXLAJK0CSFeSJ7tEo2JicHd3R1ra2tJ1oQQr41Go8Ha2hp3d3diYmKMHY4QIofKswlbUlISVlZWxg5DCJFHWFlZyfALIcQLM1rC1rt3b5ydnSlbtqz+2M2bN2nSpAklSpSgSZMm3Lp1S1/2zTff4OPjQ8mSJVm/fv0riUFa1oQQWUV+3wghXobRErbAwECCg4MNjk2YMIFGjRpx6tQpGjVqxIQJEwA4duwYixcv5ujRowQHB/Puu++SkpJijLCFEEIIIbKc0RK2unXr4uDgYHBs1apVBAQEABAQEMDKlSv1x7t06YKFhQVFixbFx8eH3bt3Z3XI4iWFhISg0WieurGtEEIIIdKXrcawRUdHU7hwYQAKFy6sH6B7+fJlPD099fU8PDz0M67ymsDAQFq1amXsMNIVGhpKo0aNcHR0xNramuLFi9O9e3diY2Nf+7UjIyPRaDSEh4c/tc68efPQaDTP/AkJCXntsT7NmDFj0Gg0NG7cOE3Zzz//jEajMRhCIIQQIu/IVgnb0yil0hx72niQWbNmUaVKFapUqcK1a9ded2jioWPHjtG8eXPKly/P1q1bOXLkCDNmzMDe3p6EhITXeu3ExMQM1evcuTNRUVH6n8aNG9OpUyeDYzVr1szUtV/1IHJXV1fCwsKIjIw0OB4UFESRIkVe6bVepYx+B0IIIV5MtkrYXFxciIqKAiAqKgpnZ2dA16J28eJFfb1Lly7h5uaW7jn69+9PeHg44eHhODk5vf6gs5lt27ZRvXp1LC0tcXFx4eOPPzb4x7R+/fq8++67jBw5EkdHR5ydnRk6dCharVZfJzo6mrfeegsrKyu8vLyYO3cuZcuWZcyYMU+97oYNGyhUqBBTp06lXLlyFCtWjKZNm/Lzzz+n+R4OHjxI9erVsba2pkqVKuzbt8+g/K+//qJcuXJYWFjg6enJ119/bZC0e3t7M2bMGHr37k2BAgXo3r07RYsWBaBq1apoNBrq16+fJkYrKytcXV31PxYWFgbHHBwc+N///oeHhwc2NjZUrVrVYIJLapfu2rVrqVatGubm5qxfv5769eszaNAghgwZgoODA05OTnz//fckJCTw3nvvUaBAAYoUKcJvv/323O+vUKFCtGzZkrlz5+qPHTp0iOPHj/P222+nqb969WoqV66MpaUlRYsWZdSoUQbft7e3N2PHjiUwMBA7Ozs8PT1ZsmQJt2/fpkuXLtja2lKiRAk2bNhgcN6M/DkaNGgQQ4cOxcnJiVq1atG7d+80rb9arZYiRYowZcqU5352IYQQT5etEra33nqL+fPnAzB//nzatGmjP7548WISEhI4d+4cp06dolq1asYMNVu6fPkyLVq0oFKlSuzfv585c+bwxx9/MGLECIN6CxcuxNTUlH///ZeffvqJadOmsWTJEn15QEAA58+fZ8uWLaxatYrff/+d8+fPP/Parq6uXLt2ja1btz43zhEjRjBhwgT27dtHoUKF6N69uz4h27t3Lx07dqR9+/YcPnyYCRMm8M033/DTTz8ZnGPKlCmUKlWK8PBwxo8frx/TGBwcTFRUFH/99VeG7tnjevXqRWhoKIsWLeLw4cMEBATQunVrDh48aFBv2LBhfPXVVxw/fpzq1asDuntqZ2fHrl27GD58OB999BFt27bF19eX8PBwAgIC6Nu3L1euXHluHH369GH+/Pn6JHrOnDl06tQJOzs7g3rr16+ne/fuDB48mKNHjxIUFMSyZcsYOXKkQb1p06ZRrVo19u3bR6dOnQgICKBbt268+eabHDhwgLp16/LOO+8QHx8PZPzP0e+//45SirCwMBYsWEC/fv309z/Vxo0buXr1Kj169MjgtyCEECJdyki6dOmiXF1dlampqXJ3d1ezZ89W169fVw0bNlQ+Pj6qYcOG6saNG/r6X331lSpWrJjy9fVVa9euzdA1Kleu/NSyY8eOpTkGxvnJjICAANWyZct0y0aOHKmKFy+uUlJS9Mfmzp2rzM3N1f3795VSStWrV0+98cYbBu9r3Lix6tOnj1JKqePHjytA7dy5U19+4cIFZWJiokaPHv3UuJKTk1VgYKAClLOzs2rVqpX67rvvVExMjL7O1q1bFaCCg4P1x7Zv364AdfHiRaWUUt26dVMNGjQwOPfo0aOVu7u7/rWXl5dq1aqVQZ1z584pQO3Zs+epMT6pZcuWKiAgQCml1OnTp5VGo1Hnz583qNOmTRs1aNAgg/iXLVtmUOfJe6rVapWjo6Nq3bq1/lhiYqIyMzNTS5cufWo8o0ePVmXKlFHJycnKzc1NbdiwQcXHx6tChQqpsLAwfXmqOnXqqLFjxxqcY8WKFcrGxkZptVqllO5edenSRV9+9+5dBaj3339ff+zJe5fRP0flypVL8xnKlCmjvvnmG/3rTp06qQ4dOjz1M+c16f3eEUKIVM/KW4zWwvbHH38QFRVFUlISly5dok+fPhQqVIjNmzdz6tQpNm/ebDCLdNSoUZw5c4YTJ07QokULY4WdrUVERFCjRg1MTB59rbVr1yYxMZHTp0/rj5UvX97gfW5ubvoJHsePH8fExIQqVaroyz09PQ26oAcOHIitra3+ByBfvnzMnTuXS5cuMXnyZIoUKcKkSZMoVaoUR48eNbje49dPPW/q9SMiIqhVq5ZB/dq1a3P58mWDyQuPx/cq7Nu3D6UUfn5+Bp/tn3/+4cyZMwZ107v2459Jo9Hg7OxMuXLl9MfMzMwoWLBghla6z5cvHwEBAQQFBbFy5UoKFSpE7dq109Tbu3cvX3/9tUG83bp14/79+1y9ejXd2GxtbbG2tjaIzcXFBTD8DjLy56hy5cppYurXr5++O/fmzZusWrWKPn36PPczCyGEeLY8u5doetKZ25CjKKWeOhnj8eNmZmZpylK731QGbsLYsWMZOnRoumXu7u706NGDHj168NVXX+Hr68ukSZOYN29eutdPjevx62fkM9jY2Dw3zszQarVoNBr27NmT5v48uSNGetdO754+6z4/T+/evSlfvjyRkZH07t37qTGPHj2ajh07pil7fNzg82J7ld9Bjx49GDZsGNu3b2f//v04OjrStGnTp31MIYQQGSQJWy7i5+fHn3/+iVar1beObN++HXNzc4oXL56hc5QuXRqtVsvevXv147MuXbpkMPbK2dlZPyHkWQoWLEjhwoW5d+9epj7D9u3bDY5t374dDw+PNGO4Hmdubg7wwgsqV6pUCaUUV69epUGDBi90jlfJx8eHqlWr8u+//z51PJ6/vz/Hjx/Hx8fnlV77Zf4cOTg40L59e4KCgti/fz+BgYHky5fvlcYnhBB5kSRsOVBsbCwHDhwwOFagQAHeffddpk2bxrvvvsuHH37I2bNnGT58OIMHD8ba2jpD5y5ZsiTNmjVj4MCBzJgxA0tLSz799FOsra2fubXOL7/8woEDB2jXrh3FixcnPj6eBQsWcPjwYT777LMMf7YhQ4ZQtWpVxowZQ7du3dizZw/fffcd48ePf+b7nJ2dsbKyYv369Xh7e2NpaYm9vX2Gr+vr60v37t0JDAzku+++w9/fn5s3bxISEkKxYsVo3759hs/1qqxbt46EhAQKFiyYbvkXX3xBq1at8PLyolOnTpiamnLkyBF2797NxIkTX/i6L/vnqF+/fjRv3pykpCSWLVv2wnEIIYR4JFvNEhUZExYWRqVKlQx+hg4diru7O+vWrWP//v1UrFiR3r1707Vr1+cmO0+aN28eHh4e1K9fn7feeovu3bvj7OyMpaXlU99TrVo14uLiGDRoEGXLlqVu3bqEhoayYMEC3nnnnQxf29/fn6VLl7J8+XLKli3L8OHD9cnCs5iamvLDDz8we/Zs3Nzc9DOMM2Pu3Ln06tWLzz77jFKlStGqVSu2bduGl5dXps/1KlhbWz81WQNo1qwZ//zzD1u3bqVatWpUq1aNCRMmvPR6bS/756h+/fr6Pz8ZbdkVQuRCWi2Ehma4ulKK0MjQDA3NeVkXL8LJk3Bt/35SMrJWaNwVuHP8tcf1LBqVFXfGSKpUqfLUle8jIiIoXbp0FkeUM12/fh03Nzf++OMPOnToYOxwRDb34MED3N3d+fHHH+nevbuxw8lW5PeOyDMuXoSAAAgJgU2boGHDZ1a/eu8qg/4ZxMrjK/mjwx90KdvltYV2/z7Urg35ruxliGtfzjgU4JOVK7F+Wq9MTBhs7wSmttB8D5gXeG2xPStvkRY2kUbq+mtnz57lv//+o3Pnzjg6OtK8eXNjhyayMa1WS1RUFKNHj8bKyirdyRBCiDzgzz+hfHnYuhUcHSE5+alVlVIsOryIMj+XYeXxleS3yI9WZWxy1ovQanV5ZPSxswwsNJjb8XF88++/fP/zz+kFB8enweYGEH8VbDxB+2p3t8kMGcMm0khKSuLzzz/n7NmzWFtbU716dbZt2/bKZ2aK3OXChQsULVoUDw8P5s6dq58IIoTII+7cgfffh9RdXVq2hDlz4OHSQU96vFUNoFnxZvza+lc87T3Trf8qjBkDm1ddY2yxAVhrYnFv3Jjwd9+l6JPDN5Lvw66+cH6x7nXpT6HCeDAxXtokCZtIo1mzZjRr1szYYYgcxtvbO0vGngghsqHt2+Gdd+D8ebCygilTYMAASGeymlKKxUcWM3jdYG4+uEl+i/xMaTqF3pV6P3Ny28tavBgmfX2f/3kPwtbkEofyF6DjxImYPdkYEXsKwtrDnSO6btA35kKRtFsDZjVJ2IQQQgjxYpKSdM1WEybo+hv9/WHhQihVKt3q0feiGfTPIFYcXwFkTasawJ490LdXEh96fISXxTHm3Ikl5MQJup4+TYUKFR5VvPQ37OwBSbGQvyTUWQH22WPcqSRsQgghhMi8Eyd0rWrh4bqWtBEjdMlbOsMhnmxVszO3Y0qzKfSp1Oe1tqoBXL4Mbdoo3nEYTXnbfwlJSGTrlSuMHj36UbKmTYHDo+Ho17rXnh3gjSAwy/9aY8sMSdiEEEIIkXFKwaxZ8MknEBcHXl66cWt16qRbPfpeNO+ufZe/InSLgDct3pRfW/9KEfuXW4IoI+LioE0bqJX8I3WdVnEuRcv8Cxdo0aIFX3zxha5Swg3Y0Q2ubgCNCVT4Rjdm7TUnkpklCZsQQgghMiYmBvr2hdWrda/feQd++gnSWRJDKcWSo0sYvHYwNx7cyNJWNd31oVcvsD/zJ+3cfiERmHH7Fm7u7vz++++6nVxu7oWwDnD/PFg4Qq0l4PrsJUiMRRI2IYQQQjzf2rW6DCgmRpegzZwJXdJfL+3JVrUmxZow+63ZWdKqlmrcODi1NoRPPMcBUHvMGH62sMDd3R0HBwc4EwR73gVtAhSqBrWX6ZbuyKYkYRNCCCHE08XFwaefQupaZfXrw/z5kM6uKkop/jz6J++tfU/fqvZd0+/o6983S1rVUi1dCvO/OcQo76GYaLQU6tQJn44d8QFISYDdA+D0LF1lnwFQ+XvIZ5Fl8b0IWThX5FohISFoNBquX79u7FAybcyYMZQtWzbPXDc98+bNw9bW9pWeMzt9PiFyhLAw3czPn38GMzP49lvdzgXpJGuXYy/T/s/2dFnehRsPbtC4WGOOvHuEfpX7ZWmydvo0vNfrJkOLvIelyQOuVahA66+/ZuPGjbrJBSEtdcmaiQVUD4JqM7N9sgaSsOVI0dHRfPjhhxQvXhyLh827LVq0YO3atcYOLdOM/Q9o/fr10Wg0aX5u376dJdePjIxEo9Gk2Ypk6NChhGZiD76slN790mg0zJw585Vep3Pnzpw9e/aVnvN1eNp3KESOFhsL774LdevqZoOWLg27dsFnn0G+fAZVtUrLjD0z8PvZj5XHV2JnbscvrX5hwzsbsrQLNNWkSdDM9hfsTW/iVLkKi06fpkiRItSpUwdO/gjRm8HSBZrugOK9sjy+FyVdojlMZGQktWrVws7Ojm+++YYKFSqg1WrZvHkzAwcO5MKFC8YOMcfp1atXmo3N7Z+2p1wWsbW1feWtS6/Sr7/+SqtWrQyOvep7ZmVlhZWV1Ss9pxAiA1avhkGDdOthmJrqlusYNQos0rZCHbt2jP6r+7Pj4g4A3ir5FtPfnI5Hfo+sjhqA6GhYu/AiXxdZDBoNiU2bsPu3Bfz8889YJl2CgyN1FavNAofKRonxRUkLWw7z7rvvopQiPDycTp06UbJkSUqXLs3gwYM5ePCgvt6FCxdo164ddnZ22NnZ0b59ey5duqQvT23ZWrx4McWLF8fOzo62bdum6T6cP38+5cqVw8LCAhcXFwIDA/Vld+7coX///jg7O2NnZ0e9evUMWhlSu7RWr16Nr68vlpaWNGjQQN9qMm/ePL788kuOHj2qb6WZN28eoGvFWbZsmUEs3t7eTJ48Wf96ypQplC9fHhsbG9zd3enbt+8LtYxZW1vj6upq8KPRaAgMDEyTlDzZIpha5/vvv8fd3Z2CBQvSq1cv4uLi9HWUUnz33XeUKFECCwsLPDw8GDFiBABFixYFoGrVqmg0GurXr5/udbRaLePGjcPT0xMLCwvKlSvHqlWr9OWprTzLly+nSZMmWFtb4+fnp+sCeCglJYU+ffpQtGhRrKysKFGiBBMnTkSrzfy+fQUKFEhzzx5PrhYsWICXlxfW1ta0atWK6dOnG3SJpNey+mQX6OOvT548iUaj4fDhwwbvmTVrFo6OjiQlJb3w55s7dy5+fn5YWlri6+vL1KlTDd6j0WiYNWsWHTt2xMbGhmLFivH777/ry5/2HQqR40RHQ+fO8NZbumStWjXYtw/Gjk2TrCUkJzAmZAwVZ1Zkx8UduNq6srTjUlZ2Xmm0ZA10E1bfKvAjpppkirZuzcwlS3B0dCSgZw/dVlMpD8CrG3i8ZbQYX5QkbDnIzZs3CQ4OZvDgwem2vhQsWBDQJQht27YlOjqaLVu2sHXrVq5cuULbtm0Ntg6KjIxkyZIlrFixgg0bNrB//35GjRqlL//ll18YMGAAvXr14tChQ6xdu5YyZcror9GyZUsuX77MmjVr2L9/P3Xr1qVhw4ZERUXpz5GQkMCXX37J3Llz2blzJykpKbRr1w6lFJ07d2bIkCGULFmSqKgooqKi6Ny5c4bvh4mJCdOmTePo0aMsWrSI3bt38/7772f6vr6ssLAwjhw5wqZNm/T38/vvv9eXjxw5knHjxjFixAiOHj3K0qVL8fTUzUTavXs3AMHBwURFRfHXX3+le43vv/+eSZMm8e2333L48GHatWtH+/btOXDggEG9UaNG8cEHH3Dw4EGqVq1Kly5duHfvHqBL+tzd3fnzzz+JiIjg66+/Zvz48cydO/eV3o9du3YRGBhI//79OXDgAK1bt3603tEL8vX1pUqVKixcuNDg+MKFC+ncuTNmZmYv9Pl+/fVXRo4cydixY4mIiOC7777j22+/5ecnNoIeO3Ysbdq04eDBg3Tu3JnevXtz/vx5IOPfoRDZllK6SQSlS+s2bre2hqlT4d9/oVy5NNV3XNhBpV8q8WXolyRpk+jn349j7x7jbb+3s3Ss2pPu34e/Z0VQy/4fMDXHpUsXNmzYwHvvvYf15QUQEwqWzlDlB6PF+FJULla5cuWnlh07diztQd0f26z/yaBdu3YpQP3111/PrLdhwwZlYmKizp07pz925swZpdFo1MaNG5VSSo0ePVpZWFio27dv6+t89dVXqnjx4vrX7u7uatiwYeleY/PmzcrGxkbFxcUZHK9QoYL69ttvlVJKzZ07VwFq+/bt+vLIyEhlYmJiEEeZMmXSnB9QS5cuNTjm5eWlJk2a9NTPvW7dOmVubq5SUlKUUkpt3bpVAeratWtPfU+9evWUmZmZsrGx0f8MGDBAKaVUQECAatmypUH9J+MNCAhQHh4eKikpSX+sb9++qlGjRkoppe7evassLCzUjBkz0r3+uXPnFKD27NnzzOu4ubmpL7/8Mk3s3bt3NzjPzJkz9eWXLl1SgAoLC3vq5x82bJg+1vSumx5AWVpaGtwzGxsbdejQIaWUUl27dlWNGzc2eE+fPn3U479u0rvO3LlzlY2NzVNfT5s2TRUpUkRptVqllFIXLlxQGo1G/fvvvy/8+Tw9PdWCBQsM3jN16lRVunRpg887fPhw/eukpCRlZWWlfvvtN6XU07/D9KT7e0cIYzp7VqkmTR79e9S0qVKP/dvxuNsPbquBqwcqxqAYg/L90VeFnAvJ2nif4YcflBpepK9a6Oen9n47USml+z1488IBpZbYKLUQpc4vM3KUz/asvEXGsOUgKoMba0dERODm5oa3t7f+WLFixXBzc+PYsWM0btwYAC8vL4NxR25ubsTExAAQExPD5cuXadSoUbrX2Lt3L3FxcTg5ORkcj4+P58yZM/rXJiYmVKtWTf/ay8srTRwvasuWLXzzzTdERERw584dUlJSSExM5OrVq7i5uWX4PJ07d2b06NH61/nzZ24rEj8/P0xNH/1VcnNzY9euXQAcO3aMhISEp97HjIiNjeXKlSvUqlXL4Hjt2rXTTDQpX768QRyA/jsFmDlzJrNnz+b8+fM8ePCApKQkvLy8Mh3TpEmTaN68ucGxIg9njUVERNC6dWuDsho1ajBnzpxMX+dxXbt2ZejQoYSFhVG3bl0WLVpEsWLFqFGjhr5OZj7ftWvXuHjxIgMGDGDQoEH648nJyWn+rj1+X01NTXFycjK4r0LkOCkp8P338L//6ZbtcHCAadN0C+Gm00q2ImIFg9cN5srdK5iamDK81nBG1R2Fpall1seejuRkWDXtX3rb/guWdpTq0xsAdzc32BoIyfehSEco0sG4gb4ESdgel8GEyFhKlCiBRqMhIiKCdu3aPbWeUuqpzdKPHzczM0tTljp253nJoVarxcXFhbCwsDRlmU14nhbnkzEkJSXpn58/f56WLVvSr18/xo4dS6FChdi3bx9du3YlMTExU9eyt7fHx8cnzXETE5NnxpDqZe5jZqT3nT557PFYUstSY1myZAkfffQRkydPpmbNmuTPn5/p06ezYsWKTMfi6uqa7j2DjH3mjN7bxzk7O9O4cWMWLlxI3bp1WbhwId27d9eXZ/bzpd6XmTNnUrNmzWde+1nfsRA5zqFDut0K9uzRve7aVZesOTunqXrl7hUGrx2s36z9DY83+LX1r5R1zl7L4yxdqqWBmgJA+YF9+eq779i9ezfrfnwbs6ubwKIQVPnJyFG+HBnDloM4ODjQrFkzfvrpJ/24pMelDrj38/Pj8uXLREZG6svOnj3LlStX8PPzy9C1XFxccHd3Z/PmzemW+/v7Ex0djYmJCT4+PgY/zo/9pddqtexJ/aWAbjLElStXKF26NADm5uakpKSkOb+Tk5PBWLjo6GiD1+Hh4SQmJjJ16lRq1KiBr68vV65cydBny6gnYwDSjBl7Hj8/PywsLJ56H80fbpKc3j1IlT9/ftzc3Ni+fbvB8e3bt2f4+0ytX716dQYPHoy/vz8+Pj4GraGvip+fH//995/BsSdfOzk5ER0dbZC0ZeTevvPOOyxdupS9e/dy+PBh3nnnHX1ZZj9f6p/xM2fOpPkz/LRkND0Z+Q6FyBbi4+Hzz6FyZV2y5uGhmxG6aFGaZE2rtMwMn0np6aVZcXwFtua2/NjiR7b32p7tkjWlYM236yhqFYHWxgWPdm2ZMWMG9rbmmB0ZpqtU+Ufd+LUcTBK2HObnn39GKUWVKlVYunQpJ06c4Pjx48yYMUPfbdO4cWMqVKhA9+7d2bt3L+Hh4XTv3h1/f38aNsz4HmmjRo1i2rRpTJ06lZMnT3LgwAG+++47/TVq1apFmzZtWLduHefOnWPnzp2MHj3aoNXN1NSUjz76iJ07d3LgwAECAgIoU6aMvjvU29ub8+fPs2/fPq5fv05CQgIADRs2ZPr06YSHh7N//34CAwOxtHzU9F6iRAm0Wi3Tpk3j3Llz/PHHH0ybNu1lb6+Bhg0bsn//foKCgjh9+jQTJ05kx44dmTqHnZ0dH374ISNGjGDu3LmcOXOG3bt3M2PGDEDXamRlZcX69euJjo7mzp076Z7n008/ZfLkyfzxxx+cPHmSL774grCwMIYMGZLhWHx9fdm3bx/r1q3j1KlTjBs37oXXert9+zZXr141+En9T8QHH3zApk2b+Oabbzh16hS//vprmlau+vXrc/PmTcaPH8+ZM2eYM2dOmlnB6WnXrh1JSUn06dOHatWqUaJEiZf6fGPGjGHixIlMnTqVEydOcOTIERYsWMA333yT4XuR0e9QCKMKC4OKFeHrr3X9h++9B0ePwhMz4QEirkVQb149Bv0ziNiEWFr7tubYu8cYXG0w+UzypT23kW3ZmEjV+7qJBJU/fo/5Cxdy+/ZtPm1yC5JiwaMNeKW/hVaO8roH0BlTpicd5BBXrlxRgwcPVkWLFlXm5uaqcOHCqnnz5mrt2rX6OufPn1dt2rRRtra2ytbWVrVt21ZdvHhRX56RQd9KKTV79mxVunRpZWZmplxcXFSvXr30ZbGxseqDDz5Q7u7uyszMTHl4eKjOnTur06dPG5xv5cqVysfHR5mbm6u6deuqU6dO6c8RHx+vOnTooAoUKKAANXfuXKWUUpcvX1bNmzdXNjY2qlixYmrZsmVpJh18//33ys3NTVlaWqqGDRuqJUuWKEA/2SKjkw7ee++9p5aPHj1aubq6qvz586tBgwapESNGpJl08LyJCSkpKeqbb75RRYsW1d+nkSNH6st//fVX5enpqUxMTFS9evWeeo6xY8cqDw8PZWZmpsqWLatWrFihL3/awHcem7yRkJCgevfurQoUKKDs7e1V79691Zdffqm8vLyeGnt6gHR/Ro0apa8TFBSkPD09laWlpWrevLn68ccf1ZO/bmbOnKmKFCmirK2tVefOndW0adOeOekgVY8ePRSgfvjhB4PjL/r5Fi1apCpVqqQsLCxUgQIFVK1atdQff/yR7j1M9eSfxfS+w/Tk5N87IofSapX6/PNHkwpKlVLqsYlgj7t+/7oaun6oMh9nrhiDcpnkov488qd+ok92pNUqNbj6b2qhn5+a+0ZrlfDggfLy8lK1KvvqJhksLahU3BVjh5lhz8pbNEpl84FbL6FKlSpPXX08IiJC3y0nXo958+YxePDgdLtvRd6ybNkyOnbs+ErH9OVE8ntHZCmtFj78ULc4Wb58MHJkugvg3k+8z7T/pjHx34nEJsQC0LdSXyY2mUhBq4LGiDzDvp9wG6t5LbDNF0ul8T+y714U77zzDqtGOvNWmRioPgeK9zZ2mBn2rLxFJh0IIYQQuU1yMvTrB/Pmgbm5bjf0twwXi01MSWT2vtmMDR1L9P1oAJoWb8r4huOp7Jb9dwHYswcO/PgzTQrEYlLsDUq91QCPe/f45auetPJaAHa+UDTA2GG+MpKwCSGEELlJYqJueY6lS3WL4K5aBY8to6RVWhYfWcz/tv6Ps7d0O89UdavKhMYTaFg04+OcjenOHRjc5QyD7RejMKH5d8PQaDTY2VjRv9wOuAeU/Ryy4Zi7FyUJm3htAgMDDbayEnnX22+/nee7Q4XIEg8eQMeO8M8/kD8/rF0LD9dwVEqx7vQ6Rm4eycFo3VaGJQuVZHyj8bQr1c6ouxRkhlK6VUnqJU0in2UKRd/ujH2JEnTv3p0u9Z1obXMGbH3Aq6uxQ32lJGETQgghcoO7d6FNG9i6FQoVgg0bwN8fgH8v/suIzSPYdn4bAB75PRhTbwwBFQMwNclZqcAvv8Cp4DDaFQkjn7UtlT4czMaNG1m0aBENCjpBTaDs/yCHfa7nyV2fRgghhMiLbt2CN9+E//6DwoVh40YoU4YjMUcYtWUUf5/4GwAHKwdG1h7Ju1XfxcrMyshBZ97BgzDkoyTGeEwEoPx7g7B0cODrr7/Go7ADPatd07WueXczcqSvniRsQgghRE4WEwPNmsGBA+DlBZs2cd7RjNErA1lwcAEKhbWZNR+/8TGf1vwUe0v7554yO7p3Dzp1gtrWf+JucRbbIkXw7daN7du3s23bNqb1dcLcFCg7Kte1roEkbEIIIUTOdfmybkLB8ePg68u11Uv4+vRPzFg8g8SURExNTBlQeQCf1/0cV1tXY0f7wpSCQYPgypnbfFZSt8WU/6efks/cnPHjx+PoYEffGtfAthh4v/Ocs+VMkrAJIYQQOdHZs7pk7dw57lbyY8q4N5m8vA73EnVrX3Yr142x9cdS3KG4kQN9efPnw++/Q2+Pn7EiFpc33sC9QQOUUnTv2pWOvvuxsbwLZXJn6xpIwiaEEELkPFeuQJ06cOUKW1qUokv9GK6FTwbgzRJv8nXDr6noWtG4Mb4ip07pdtJytzhNQ/vFaDQmVB42TD+rtXsdU8h3FWyKQtEeRo729ZG9REUaGo0mQ/s65pbr5ha2trbMmzcvw/XnzZuHra3t6wtICPH6fPUVXLnC7hbleav2Ra49uE4NjxqEBobyT7d/ck2yBjB6NCTEJTGy7Eg0KgWfjh0p4OvLvHnzGDvmC7T7R+oqlh0FJmbGDfY1koQthwkMDESj0fDVV18ZHA8JCUGj0XD9+vVMnatVOhv/RkVF0bp165eO9VUKDg5Go9Fw6dIlg+PFixenQIECpKSk6I/FxcVhbm5OUFBQhs79uhPFMWPGoNFo9BveP+7nn39Go9FQtmzZ13Z9IUQuc/EizJ5NhBO8WecC95Pu8075d9jeezt1veoaO7pX6vhxWLwYOrjMpMCDo9i4u1Pxk0/Yu3cvAwcOJDT4D7T3IsG+bK7a1SA9krDlQJaWlkycOJFr1669lvO7urpi8cRec8ZWp04dzMzMCAkJ0R+7ePEily5dwtLSkv379+uP79ixg6SkJBo0aJClMSYmJj61zNXVlbCwMCIjIw2OBwUFUaRIkdccmRAiVxk/ngvWSTTtb82NxNu0LNGSoLeCMNHkvn/Sx4+HYpaHeKvQr6DRUGP8eG4/eED79u1xcXZicb+rmOYD/L/LtWPXUuW+bzcPaNCgAd7e3owbN+6pdVJSUujTpw9FixbFysqKEiVKMHHiRLRaLaBr9Zk/fz7//PMPGo0GjUajT4aebHE6fPgwjRs3xsrKCgcHBwIDA7lz546+PLWl7vvvv8fd3Z2CBQvSq1cv4uLi9HWCg4OpU6cOBQsWxMHBgWbNmhEREZHhz2xjY0PVqlXZunWr/tjWrVupVq0ajRo1SnPcy8uLokWLsmfPHpo2bYqjoyP58+endu3a7Ny5U1/X29sbgI4dO6LRaPSvAVavXk3lypWxtLSkaNGijBo1yiAp8/b2ZsyYMfTu3ZsCBQrQvXv3p8ZfqFAhWrZsydy5c/XHDh06xPHjx3n77bfT1P/ll1/w8fHB3NwcHx8ffv31V4Py06dPU79+fSwtLSlZsiRr1qxJc47Lly/TpUsXChYsSMGCBWnZsiWnTp16aoxCiBzg/HmuL5pN0x5wySyOWp61+LPjn5jly31dgadPw7JFcbzrPhwNKZQODMShYkW6du1KdHQ0y8e9gZPVPSjcAgo3NXa4r50kbDmQiYkJEyZMYObMmZw5cybdOlqtFnd3d/78808iIiL4+uuvGT9+vD5hGDp0KJ06daJx48ZERUURFRVFzZo105wnLi6O5s2bY2try+7du1mxYgX//vsvvXv3NqgXFhbGkSNH2LRpE0uWLGHFihV8//33+vL79+/z0UcfsXv3bkJCQrC3t6d169bPbJV6UoMGDdIkZvXr16d+/fppjjdsqNsP7+7du/To0YOwsDB2795NxYoVefPNN/Vdx3v27AHg119/JSoqSv96/fr1dO/encGDB3P06FGCgoJYtmwZI0eONIhpypQplCpVivDwcMaPH//M+Pv06cP8+fP1SfOcOXPo1KkTdnZ2BvVWrFjB4MGD+eijjzhy5Agffvgh7777LqtXrwZ03227du3QarXs3LmToKAgxowZQ0JCgv4ccXFxNGjQAEtLS0JDQ9m5cyeFCxemcePGBom0ECJnuTvhS97snMwJRyjnXI413dZgbWZt7LBei/HjoZPTFFzNz2NfogTl33+fXbt2ERoayozv/kcV8xWgyQf+k40datZQuVjlypWfWnbs2LE0xxiDUX4yIyAgQLVs2VIppVT9+vVV586dlVJKbd26VQHq2rVrT33vsGHDVKNGjdI9l8F9ALV06VKllFKzZs1S+fPnV7Gxsfry1GudOnVKfx4PDw+VlJSkr9O3b1+Daz3p3r17ysTERIWFhaV73fRs2rRJASoyMlIppZSXl5fatGmTOnHihLK1tVVJSUnq7t27ytTUVP3222/pnkOr1SpXV1eD8vSuW6dOHTV27FiDYytWrFA2NjZKq9Xqr9+qVaunxptq9OjRqkyZMio5OVm5ubmpDRs2qPj4eFWoUCEVFhamL09Vs2ZN1atXL4NzBAQEqFq1aimllFq/fr0yMTFR58+f15eHhYUpQM2dO1cppdScOXOUj4+PPlallEpOTlYODg5qyZIlSiml5s6dq2xsbJ4bv3h10vu9I0RGxZ85oRoFaBRjUEUneagrsVeMHdJrc/asUhXstquFfn5qUfny6mZEhL7s1KlTSm19U6mFKLV7kBGjfPWelbdIC1sONnHiRJYuXUp4eHi65TNnzqRKlSo4OTlha2vL1KlTuXDhQqauERERQfny5Q1agWrWrImJiQnHjh3TH/Pz88PU9NH4ATc3N2JiYvSvz5w5Q7du3ShevDj58+fHxcUFrVabqXhq1qyJhYUFW7du5dy5c/pWQV9fX+zs7AgPDycsLIzk5GT9+LWYmBgGDBiAr68v9vb22NnZERMT89zr7t27l6+//hpbW1v9T7du3bh//z5Xr17V16tSpUqG48+XLx8BAQEEBQWxcuVKChUqRO3atdPUi4iIoNbDzZpT1a5dW3+/IyIicHd3Nxj7Vr16dUxMHv113rt3L+fOncPOzk4fv729Pbdu3Xpqq6wQIvtK0abQfVZzNhdVuCRbsrF3CIXtChs7rNdm0tjb9HX9HIDygwdzJSWFlStXAuBjcxaurAWz/FBujPGCzGK5e4ReJqnRytghZErVqlXp0KEDw4YN43//+59B2ZIlS/joo4+YPHkyNWvWJH/+/EyfPp0VK1Zk6hpKKf1aN096/LiZmVmastSuP4DWrVvj7u7OL7/8gru7O6ampvj5+WWqS9TKyoo33niDkJAQlFJUq1YNKyvdXnj16tUjJCSEW7du4evri7u7OwABAQFER0czdepUvL29sbCwoFGjRs+9rlarZfTo0XTs2DFNmZOTk/65jY1NhuMH6N27N+XLlycyMjJNt/Lj0rvnqceUev6fU61WS8WKFVm8eHGaMgcHh0xELIQwNqUU7y7pwXKrc9jHw/o2i3PFYrhPc/48mG7+Cge7GGxKVcStQweqv/EGd+/epXHDBtjuG6KrWGYUWDobN9gsJAlbDjd+/Hj8/PwIDg42OL59+3aqV6/O4MGD9ceebFkxNzc3WA4jPX5+fgQFBXH37l19K9u///6LVquldOnSGYrxxo0bREREMH36dH3L1759+0hOTs7Q+x/XoEEDgoKC0Gq11K9fX3+8fv36/PXXX9y6dUs/fg109+GHH36gZcuWAERHRxMVFWVwTjMzszT3wd/fn+PHj+Pj45PpGJ/Fx8eHqlWr8u+///LXX3+lW6d06dJs377dIKHbvn07fn5+gO47uXz5MhcvXsTT0xOA3bt3GyTI/v7+/PHHHzg6OlKgQIFX+hmEEFnr8y2fM+vkH1gmwerrTanwRhtjh/RazflkLW/YrSPZxIr6331Nz169OHfuHFu2bME25k+4cwRsvKHkB8YONUtJl2gO5+PjQ//+/Q0G+AP4+vqyb98+1q1bx6lTpxg3bhyhoaEGdby9vTly5AgnTpzg+vXrJCUlpTl/9+7dsbGxoWfPnhw+fJht27YxYMAA2rdvn+FkpmDBgjg6OvLrr79y+vRpQkNDGThwoEEXakY1aNCACxcusHLlSoOErV69eoSFhbFv3z6D5Tx8fX35/fffOXbsGHv27KFLly6Ym5unuQ+bN2/m6tWr3Lp1C4AvvviCRYsW8cUXX3DkyBGOHz/OsmXL+OyzzzId85PWrVtHTEwMhQun353x6aef8ttvvzF9+nROnTrFjz/+yMKFC/XXbty4MaVKlaJnz54cOHCAnTt38vHHHxvcz+7du+Pi4kKbNm0IDQ3l3LlzbNu2jSFDhshMUSFykOm7pzN++3jyaeHP5RrqfPqjsUN6rc4cvIbHUd0KCEX7fcZPCxfy999/M2XKFOq8UQEO6bpJqfgt5LM0YqRZTxK2XOCLL75Ik/wMGDCATp060a1bN6pWrUpkZCRDhgwxqNOvXz9Kly6tH+e2Y8eONOe2trZm/fr1xMbGUq1aNdq0aUONGjUyvCgt6Ga1LlmyhEOHDlG2bFnee+89xo0b90Jrvb3xxhtYWVmRkJBgMKu1VKlS2Nvbo9VqDRK2oKAg7t27R+XKlenSpQu9e/c2WLoD4LvvvmPr1q14enpSqVIlAJo1a8Y///yjXzqkWrVqTJgw4ZWsmWZtbU3BggWfWt62bVt+/PFHpk6dip+fH99//z0///yzfjFjExMTVqxYgVarpXr16vTs2ZPPP//c4H5aW1uzbds2ihUrRseOHSlVqhQBAQHcunXrmdcWQmQfp2+eZsgG3e/toFXQuloP8PU1clSv14phc7HNF0u0bW3KvNOYr776io4dO+p6iw5/CfEx4FgDiqQdrpLbaVRGBsTkUFWqVHnqgPyIiIgMd+kJIcSrIL93REYppWj6W1M2ndtEzwMw/2gJ2L0bcvEQh/MnYtnUthFWJnH4fLOUSi1KsGbNGkqXLk0p12RYVxFQ0GwPOPgbO9zX4ll5i4xhE0IIIbKZRYcXsencJhziYPIOGwhZmauTNYA/Pv2TIiZxxFhXp9tbujG77dq1A6Vgc31QKVDivVybrD2PdIkKIYQQ2cjNBzf5ePV7AEzeAE6/LoSHk45yq0uRieQ/8TsA1T7oxd69e/nyyy+5ffs2RC6EmG1g4QQVnr7DT24nCZsQQgiRjQxbPpBryXeoGwmBbUdDm9w9KxRg3qf/UMD0GncsfKnxTm2mT5/OpEmTyJdyD/YP1VWqNBHM8+4YXEnYhBBCiGwiLGI9s88sxSwFfnnQCM3/vjB2SK/dlctaLA7otk306xVIbGwsS5YsoVu3bthFTob4aHCsCUV7GjlS45KETQghhMgGEpPiGbBAN/txxHFHSv26Akxy/z/Tvw4Pw938DHGmLtQe0IJFixYRFxdHv8714OSPoDGBqtN1j3lY3v70QgghRDYxaXxLIizvUuKWCSPGhcBjWwLmVlevQsoOXeuad8d3MDEzY9asWVSsWJEqKT+D0kKJwVCwonEDzQZklqgQQghhZKf/nMm45C1gCjNrfI1lyTLGDilLTP/8MKWt9pBkYkPdDzsSFxeHt7c3LaoUQHN9Hli6QPkvjR1mtiAJmxBCCGFEKiKCQesGk+ANPfP507DzcGOHlCViYuDexnlgCy7NO2FuZ4c5sGLJXFjtCwlApUlgXsC4gWYTkrAJIYQQRvTXV93Z5JuCQ7I5k4esM3Y4WWba+Jv422xESz7qDnmHu3fvcvXqVUokLoeEa+BUC7zfMXaY2YaMYRMAaDQali1bluH6ISEhaDQarl+//hqjeiQ8PByNRkNkZGSWXC+rBQYG0qpVq5euI4TIWeJDNjHUaT8AXzX6CidbZyNHlDVu3oTDS9aRT5OCbYXaWLu68ttvv+Hr68vxrdN0lcp8DhqNUePMTiRhy0FmzpyJjY0NiYmJ+mOJiYlYW1tTrlw5g7qnTp1Co9GwZcuWDJ07KipKv1flqzJv3jxsbW1f6TmfpX79+mg0mjQ/Xbp0ybIYhBAiw5Riyuw+RBaEcjjTr+7Hxo4oy/zwA1S1XANAxXdaoZTi559/pkr54pRyiAY7Xyjc1MhRZi+SsOUgDRs2JC4ujt27d+uP7dq1C3t7e06ePMm1a9f0x0NCQrCwsDDYIP1ZXF1dX2gz9uymV69eREVFGfz88ssvxg5LCCHSuLzqN8Z7XQBg2ttzMDXJG6OUYmNh0U/n8bE+hMbCGo8GDQgLC+Po0aO82+RhWuL7fp5fxuNJcjdyEF9fX9zc3Ni6dav+2NatW2ncuDFVqlQhJCTE4HiNGjWwtLREKcXEiRMpXrw4VlZWlCtXjt9//93g3E92ie7atQt/f38sLS2pVKkSa9euRaPRGFwD4ODBg1SvXh1ra2uqVKnCvn37AF3C2KtXL+7fv69v6RozZgygaxUcNmwYHh4e2NjYULVqVdavX29w3uDgYEqVKoWlpSV16tTh5MmTGbpH1tbWuLq6GvzY29sDEBkZiUajYfny5TRp0gRra2v8/PzYuHGj/v1JSUl88MEHuLm5YWFhgaenJ8OHPxoA/LzYU7uK161bR+XKlbGysqJOnTpcunSJ0NBQKlSogK2tLa1ateLGjRtp4v/qq69wcXHB1taWXr168eDBg6d+1ox8r0KIbEqrZcTqj7hvDu1My9GwTN4Z7vDzz1DORNe65t2sMaZWVkyfPp2CBfLTucwpMLWDYgFGjjIbUrlY5cqVn1p27NixLIzk1enevbtq0KCB/nX9+vXV7Nmz1ciRI9WgQYP0xwsXLqy+/PJLpZRSI0eOVL6+vmrdunXq7NmzauHChcra2lqtWbNGXx9QS5cuVUopdffuXeXo6Ki6du2qjhw5ojZs2KD8/PwUoLZu3aqUUmrr1q0KUFWrVlVbtmxRERERqmnTpqpUqVJKq9WqhIQENW3aNGVtba2ioqJUVFSUunv3rlJKqW7duqnq1aur0NBQdebMGfXjjz8qMzMzdeDAAaWUUhcuXFAWFhZq8ODBKiIiQi1ZskS5u7srQJ07d+6p96ZevXrqvffee2r5uXPnFKBKliyp/v77b3Xy5EnVs2dP5eDgoI9t8uTJysPDQ4WGhqrz58+rHTt2qKCgIP05nhf74/dl27Zt6uDBg6pMmTKqZs2aqmHDhuq///5Te/bsUd7e3mrw4MH68wYEBChbW1v19ttvq8OHD6vg4GDl5uam3n//fYM6LVu21L/OyPcqspec+ntHvHr/BY1TjEGZ/w91Jirv/Lm4d08pR0et+s6nmVro56eu7Nih7t69q2xtbdUnXcsotRClwj80dphG86y8RRK2xy3EOD+ZMHv2bGVpaani4+PVgwcPlKWlpTp9+rRav369KlWqlFJKqYiICAWosLAwde/ePWVpaam2bdtmcJ4PP/xQtWjRQv/68YRt5syZqmDBgiouLu7RrVm4MN2ELTg4WF9n+/btClAXL15USik1d+5cZWNjY3Dd06dPK41Go86fP29wvE2bNvqEc8SIEapEiRJKq9Xqy8eNG5ehhM3MzEzZ2NgY/EyfPl0p9Shhmzlzpv49ly5d0t8rpZR6//33VcOGDQ2unZnY07svP/74owLU3r179cdGjx6typQpo38dEBCg7O3t9YmjUkr99ttvytzcXN27d09fJzVhy+j3KrIXSdiEUkqlJMSr6u9ZKMaghk9ubuxwstTUqUoVtzqoFvr5qeV166qU5GSllFLXLkao6JnmSi3UKBV7yrhBGtGz8pa80WGeizRo0ID4+Hh27tyJUgpHR0eKFy+Oq6srZ86c4erVq2zduhVra2uqV6/OgQMHiI+Pp3nz5mgem22TlJSEt7d3utc4fvw4ZcuWxcrKSn+sevXq6dYtX768/rmbmxsAMTExeHh4pFt/3759KKXw8/MzOJ6QkEDDhg0BiIiI4I033jCIt0aNGs+4K4907tyZ0aNHGxxzcnLKUMygm4nZpEkTfH19adq0KW+++SYtWrTAxMQkQ7Gndw0XFxcAg4khLi4u+ms+/p7HJ2nUqFGDxMREzpw5Y3A+gGPHjmX6exVCZA+LZrzLLqcEXB/kY+TQRcYOJ8vEx8OkSdDYfjUA3i1bonm49Zbjnb/ALhHcWoKdjzHDzLYkYXtcN2XsCJ6rWLFieHl5ERISglKK+vXrA2BjY0PlypUJCQkhJCSE2rVrY2ZmhlarBWD16tUUKVLE4FxmZmbpXkMpZZAEPMvj50h9T+o106PVatFoNOzZsyfN9VMTRKVe/Huwt7fHx+fZf9mfFbO/vz+RkZEEBwezZcsWAgICqFChAhs3bsxQ7M+6xpPHnnWfnudFvlchhPHdu3ONYZfngQ18U7QfdjYFjR1Slpk3D6KvJFHHT7fWnHfr1qxcuZJJkyaytHck7tZAyQ+MGmN2JglbDtSgQQO2bt2KUoqAgEcDM+vXr8+WLVsICQnhk08+AcDPzw8LCwvOnz+fphXoaUqXLs2CBQt48OCBPhF5fGZqRpmbm5OSkmJwrFKlSiiluHr1Kg0aNEj3fX5+fixfvtwgcfzvv/8yff0XZWdnR8eOHenYsSOBgYG88cYbnD59OkOxv4zDhw9z//59bGxsAN1nNjc3p3jx4mnqvsj3KoQwvm9+7soVGy1VblnR8/MfjR1OlklKggkToJztv1hzC/vixSlYqhQzPvyQi5GncLG4AflLgWsTY4eabUnClgM1aNCARYt0zehz587VH69Xrx6dOnXi7t27+oTCzs6OoUOHMnToUJRS1K1bl3v37vHff/9hYmJC//7905y/e/fufP755/Tr14+RI0dy5coVxo8fD5DhljcAb29v4uPj2bhxI5UqVcLa2hpfX1+6d+9OYGAg3333Hf7+/ty8eZOQkBCKFStG+/btGThwIN999x0fffQR7777LocPH2bmzJkZumZcXBxXr141OGZubo6Dg0OG3j9lyhQKFy5MxYoVMTMzY9GiReTPnx8PDw+sra2fG/vLSE5Opnfv3nzxxRdcuXKF4cOH069fP30C97gX+V6FEMa14thyvknYDBqY9sYYTPLlnX+Cly6F8+ehb+nfAF3r2tmzZ9m4cSNje5fCNN8NKPGuLJT7LK95/JxR5cZZokrpZlECysPDw+D43bt3lampqcqfP79KfjiQUymltFqt+uGHH1Tp0qWVubm5cnR0VI0bN1YbNmzQ1+GxSQdKKbVz505VsWJFZW5uripWrKiWLVumAPXff/8ppR4Nrr927Zr+PamD+vfs2aM/NnDgQFWoUCEFqNGjRyullEpMTFSjR49WRYsWVWZmZsrFxUW1bt1ahYeH69+3Zs0a5evrqywsLFTNmjXV77//nqFJB0Can1q1aj01vic/+6xZs1SlSpWUra2tsrOzU3Xr1lU7duzQ131e7Ondl6VLl6on/6rNmDFDFSpUSP86dULBl19+qZycnJSNjY3q2bOnun//fpo6qTLyvYrsJSf/3hEvJ+x8mLIYa64YgxrbKr9SKSnGDinLaLVKVamiVGnrXWqhn5/6s1o1FX/rlvr888+ViYmJujirkG4CXh6ebJDqWXmLRqmXGDCUzVWpUoXw8PB0yyIiIihdunQWR5RzrVq1inbt2hETE4Ojo6OxwxEiR5LfO3nT0Zij1J5bm9vxtxm4B34u+TGa76YYO6wss2MH1K6tGOfzDsXMD1Bu8GD8+vfH29ubcqWLszYwFCydod3VPN/C9qy8Je+0x4pMmT9/PsWKFcPT05MjR47w0Ucf0bp1a0nWhBAiEy7FXqL5wubcjr9N23MW/LQ2Ac0XeWu7vGnToKLtNoqZH8CiYEFK9exJSkoKo0aNwtc2EggFx5p5Pll7HtnpQKQrOjqaHj16ULJkSd577z1atGghq+gLIUQm3HpwixYLW3Ap9hK18pdl0cIE8nkXhapVjR1alomMhBV/aenk/AMAfn37YmZjg7m5OQMHDqSh721dRceMbaOYl0nCJtL12WefERkZSUJCAufPn+fnn3/Gzs7O2GEJIUSOEJ8cT9slbTkSc4TSjqX5+3RVrJKBTp3yVEvSTz9BZZuNeFkex8rZmRJdunDjxg1mzpxJbGwsXNuhq+gkCdvzSMImhBBCvEIp2hTe+esdtp3fhrudO8Fd1uCwVLd3Jp07Gze4LHT3LgTNTqajs275krIDB2JqacmiRYsYNGgQ504ehDtHwcQcHCobOdrsTxI2IYQQ4hVRSvFh8Icsj1iOvYU967qvo8j+s3DtGpQoARUrGjvELDNvHpRjNW4W57Dx8KBYu3YABAUFUblyZSq4xQFKl6zlszRqrDmBJGxCCCHEKzJh+wSm75mOeT5zVnVZRTmXcrBkia6wc+c80x2q1cJP3yfSzmkGAOXfe4985ubs37+fAwcO0Lt3b7j+sDtUxq9liCRsQgghxCuw4OACRm4ZiQYNC9svpJ53Pd0S/3/9pauQh7pD//kHitxcjrP5ZfIXK45Xy5aArnXNwsKCrl27wrV/dZWdahkx0pxDEjYhhBDiJUXdjWLgmoEAfN/8e972e1tXsH493LwJfn5QtqwRI8xaM35WvFloPgDl3x+MSb58AFy4cIH27dtT0N4ObuzSVXasYawwcxRZh00IIYR4SV9t+4oHyQ9oW6ot71d//1FBUJDusWdP4wRmBDExcCbsEO94XcTC0RmPRo30ZatWreLBgwdw+zAk3wPbYmDlasRocw5pYRM5RqtWrQgMDNS/rl+/PoMHD87yOK5fv45GoyEkJCTLr/2ycYwZM4ayeeh/+UJkhbO3zjJr3yxMNCZ83fDrRwUxMbB6NeTLl6cStj//hDds/wGgaMsWmOTLR1JSkn6fZysrq0fLeThKd2hGScKWwwQGBqLRaPQ/jo6OtGrViuPHj7+ya0RGRqLRaJ66PcbjHo/Fzs6OKlWq8FfqeI3X7K+//uKbb77JUN2QkBA0Gg3Xr19/zVHppH5Pffv2TVP22WefodFoaNWqVZbEIoR4vUaHjCZZm0yP8j3wc/J7VLBwISQnQ4sWULiw8QLMYot+T+YN+2AAvB+OXfvzzz/x8vLi0KFDukqpEw5k/FqGScKWAzVu3JioqCiioqLYsGEDDx48oN3D6dLG8OuvvxIVFcWePXuoUKECHTt2ZOfOnenWTUxMfGXXdXBwyNaL+Xp6erJkyRLu37+vP5acnMxvv/1GkSJFjBiZEOJVORx9mIWHFmJmYsaY+mMeFSgFc+bonvfubZTYjOHMGbh3ZBf2pjewKeJNQT8/lFJMnDgRHx+fRy381yRhyyxJ2HIgCwsLXF1dcXV1xd/fn48//pjjx4/rxgU8dPnyZbp06ULBggUpWLAgLVu25NSpU/ryixcv0qZNGxwcHLC2tqZUqVIsXrwYgKJFiwJQtWpVNBoN9evXf2Y8BQoUwNXVlVKlSjFz5kwsLS35+++/AfD29mbMmDH07t2bAgUK0L17dwD+/fdf6tWrh7W1Ne7u7gwaNEi36vVDcXFxBAYGYmtri4uLC+PHj09z3Se7RBMTExk5ciReXl5YWFhQrFgxfvjhByIjI2nQoAEATk5OaDQafddq6i+S4sWLY2VlRbly5dJswbVnzx4qV66MpaUllSpVYteuXc+8H6nKly9PiRIl+PPPP/XH/vnnHywtLdPcU61Wy7hx4/D09MTCwoJy5cqxatWqTMdx7NgxWrZsiZ2dHc7OznTt2lXfDSGEePVGbRmFQjGwykC8C3g/KggPh6NHwckJHrYy5QWLFkFNe113aLHWLdFoNGzYsIFDhw7x6aefYmJiAvcvQtxFMCsA9n7PPqHQk4Qth7t79y5LliyhXLlyunEB6JKdBg0aYGlpSWhoKDt37qRw4cI0btyYuLg4AN59913i4uLYunUrR48eZdq0aRQoUACA3bt3AxAcHExUVFSmujjNzMwwNTUlKSlJf2zKlCmUKlWK8PBwxo8fz+HDh2natClvvfUWBw8e5K+//nq0Ls9DQ4cOZePGjSxfvpzNmzezf/9+tm3b9sxrBwQEsGDBAqZMmUJERARz5syhQIECeHp6snz5cgCOHj1KVFQU33//PQCff/45c+bMYfr06Rw7dowRI0YwYMAA/vlH9wvn/v37tGzZkmLFihEeHs6ECRMYOnRohu9Hnz59CEoddIxuSnuvXr3QPLEW0/fff8+kSZP49ttvOXz4MO3ataN9+/YcOHAgw3FERUVRt25dypYty+7du9m0aRP37t3jrbfeQqvVZjhmIUTG/HvxX1afXI2NmQ2j6owyLEz9e9+jB5ibZ31wRqAULFkYT1W7TQB4vfkmAJMmTcLNzY1u3brpKurHr9UAjaQhGSWzRB+zqEwZo1y329GjmaofHByMra0toPuH3NPTk7Vr1+rLFy9ejFKKuXPn6hODX375BWdnZ9asWUOnTp04f/48HTp0oEKFCsCjVjXQtUIBFCpUCFfXjM/eSUhIYNKkScTGxtLosVlB9erV47PPPtO/7tmzJ507d2bIkCH6YzNmzKBSpUrExMRgbW3NnDlzCAoKolmzZgDMnTsXDw+Pp1771KlTLF68mHXr1tG8eXMAihUrpi93cHAAwNnZGUdHR/29mzJlChs2bKBOnTr6+7B7926mT59Oy5YtWbhwIYmJicydOxdbW1vKli3LqFGj6NGjR4buSbdu3Rg6dCinTp3Czs6O4OBgfvzxR7744guDepMnT2bo0KH6X2hjx45l27ZtTJ48md9//z1DccyYMYMKFSrw7bff6o8tWLAABwcHwsPDqVatWoZiFkI8n1KKkZtHAvDxGx/jYuvyqPDBA/jjD93zXr2MEJ1x7NsHNldCsfK8T8EyZcjv7c25c+fYunUrEyZMwDw1cb0u+4e+CEnYcqC6desya9YsAG7evMnPP/9M06ZN2bVrF56enuzdu5dz586lGd8VFxfHmTNnAPjwww8ZOHAgwcHBNGrUiHbt2lG58ovt5dajRw8CAwN58OAB9vb2TJ48mRYtWujLq1SpYlB/7969nD59miWpq3+j++UHcObMGaytrUlMTKRGjUdr89ja2lKuXLmnxrB//35MTEz0XZ8ZcezYMeLj42nevLlBi1dSUhLe3t4AREREUL58eX2CDBjE9TwFCxakXbt2BAUFUaBAAerXr59m/FpsbCxXrlyhVi3DsRy1a9fWJ+IZiWPv3r1s27bNoE6qM2fOSMImxCu04cwGQs+H4mDlwNCaT7S6r1gBd+5AtWp5au2133+Hmva631lFH3YDFy1alJMnT+obAgAZv/aCJGF7TGZbuozF2toaHx8f/evKlStjb2/PrFmzGDduHFqtlooVK+rHpD0utaWpT58+NGvWjLVr17Jp0yZq1qzJiBEjGDNmTKbjmTRpEs2bNyd//vw4OzunKbexsTF4rdVq6du3Lx9//HGauu7u7pw4cSLTMaQmfJmR2k24evXqNEmUmZnZC5/3Sb179yYgIABbW1vGjh371HpPdpM+fiwjcWi1Wlq2bMnkyZPTlLm4uKTzDiHEi9AqLSO36FrXhtcajr2lvWGF1O7QPDTZICUFVi6JZYx9KGg0eL35JlqtFhMTE4oXL/6oYtI9uH0QNPmgkPwnMjOk8zgX0Gg0mJiY6Men+fv7c/r0aRwdHfHx8TH4SU3YADw8POjfvz9//vknY8eO1bfapTZbp6SkZOj6rq6u+Pj4pJuspcff35+jR4+mic3HxwcrKyt8fHwwMzPjv//+07/n/v37HDly5Jnn1Gq1bN26Nd3y9D6Tn58fFhYWnD9/Pk0cXl5e+jqHDx82mOn5eFwZ0ahRI8zNzbl+/Tpt27ZNU54/f37c3NzYvn27wfHt27fj5+eX4ThS76uXl1eaz5OdZ9MKkdMsPbqUfVH7cLNzY3C1J9aC3LoVNm8GS0vo0sU4ARpBSAh4xG3FzCQJl2rVsHJyYtSoUbz99tuG/+G8tgOUFgpWAlObp55PpCUJWw6UkJDA1atXuXr1KhEREbz//vvcu3eP1q1bA9C9e3dcXFxo06YNoaGhnDt3jm3btjFkyBD9TNEPP/yQ4OBgzp49y4EDBwgODtYnB87OzlhZWbF+/Xqio6O5c+fOK41/2LBh7N69m4EDB7J//35Onz7NmjVrGDBgAKDr/uzTpw/Dhg1j48aNHD16lN69ez8zgSxRogSdOnWib9++LF++nHPnzhEWFsZvv/0GgJeXFxqNhn/++Ydr165x79497OzsGDp0KEOHDiUoKIjTp09z4MABZs6cqU9eu3XrhqmpKb179+bo0aNs3LiRr7/++qlxpEej0XDo0CHOnTuHhYVFunU+/fRTJk+ezB9//MHJkyf54osvCAsL04/zy0gc7733Hnfu3KFz587s2rWLs2fPsmnTJvr378/du3czFbMQIn2JKYn61rUx9cZgZWb1qPDmTd0kA4Dhw8HePp0z5E5LlkAl21AA/c4Ga9eu5datW4a9B1cejrd2bZLVIeZ4krDlQJs2baJw4cIULlyY6tWrs2fPHpYuXapfKsLa2ppt27ZRrFgxOnbsSKlSpQgICODWrVsULFgQ0HWfvf/++/j5+dGkSRNcXFyYP1+375upqSk//PADs2fPxs3NjTZt2rzS+MuXL8+2bduIjIykXr16VKhQgREjRhh0202ePJkGDRrQrl07GjRoQNmyZalbt+4zz7tgwQK6devGBx98QKlSpQgMDNQnm+7u7nz55ZeMGjUKFxcX/XIg48aNY8yYMUyePJkyZcrQpEkTli9frp+EYWtry5o1azh16hT+/v4MHTrUYFB/RtnZ2ZE/f/6nln/wwQd8+umnfPbZZ5QtW5YVK1awfPlyKlasmOE43Nzc2LFjByYmJjRv3pwyZcrw3nvvYWFh8dREUQiROb+E/8LZW2cp5ViKXpUem1CgFPTvD5cvQ82aMGrU00+SyyQlwYrlSZS31Y1Nc69Xj6ioKA4dOkTTpk0fVVQKruhm4OOed5Y6eVU06lUM0smmqlSp8tTV+iMiIihdunQWRySEyMvk907OFpsQS/EfinM97jqruqzirZJvPSoMCoI+fcDODg4ehMdm3ud2wcHwSYfdfO7di/zFitFq9Wp+++03evbsyd69e/H399dVjD0Ba0qBuQO0jwGTfMYNPBt6Vt4iLWxCCCFEBkzcMZHrcdepXaQ2rX1bPyo4dQo++ED3/Oef81SyBrru0Iq2unUy3evVA2DDhg04OjrqewkAuPywda1wc0nWXoDMEhVCCCGe48rdK0zZOQWAiY0nPhqXlZQE3brB/fu6x3feMWKUWS8hQbeKyShH3fg1t4dDV6pXr07JkiV1Oxukku7QlyIJmxBCCPEco7eO5kHyAzqU7kANz8fWQBw9WrcNlZeXrnUtj9mwAczjLuFucRYzOzucKlUCMNg2EICkWIjZptvZoHBzI0Sa82XLLtGpU6dSpkwZypYtS9euXYmPj+fmzZs0adKEEiVK0KRJE27dumXsMIUQQuQBx64dI+hAEPk0+Rjf6LF9jUNCYMIEMDHRrRqbh2aFptJ1h+pa1wrXrImJmRlnz55NOzM9aiOoZHCsCRYO6ZxJPE+2S9guX77MDz/8QHh4OEeOHCElJYXFixczYcIEGjVqxKlTp2jUqBETJkwwdqhCCCHygBGbR6BVWvpX7o9vIV/dwVu3dEt4KKWbEVq7tnGDNIIHD2DVKqhkpxu/ltodOmDAAP12f3qp3aFu0h36orJdwgaQnJzMgwcPSE5OJi4uDjc3N1atWkVAQACg2+R75cqVxg1SCCFErhd2Poy/T/yNjZkNo+uN1h1UCgYMgEuXoHp1+N//jBukkQQHQ9L9OMrY7gaNBrc6dYiLiyMsLIyGDRs+qqi0j9Zfk/FrLyzbJWzu7u4MHTqUIkWKULhwYezt7WnatCnR0dEULlwYgMKFCxMTE2PkSIUQQuRmSSlJfBCsm/35ac1PH23wvno1LF0KtrawcCE83Mour1myBCrYhmFKIoXKlsWyUCFCQ0NJSEigSZPHFsa9tR/io8HaE+zzzt6qr1q2S9hu3brFqlWrOHfuHFeuXOH+/fv8/vvvGX7/rFmzqFKlClWqVOHatWuvMVIhhBC52Tfbv+HA1QN4F/BmSE3driNotfD557rnX30Fj++TmYc8eABr1kAThz8A8G7VCoA///yT/Pnz06BBg0eVLz/WHZrOnskiY7JdwrZp0yaKFi2Kk5MTZmZmtG/fnn///RcXFxeioqIAiIqKeuq+lf379yc8PJzw8HCcnJyyMnQhhBC5xMGrBxm3bRwAQW8FYWtuqytYsgQOHwZPTxg40IgRGtf69VAo+Th+NnswtbGhWNu2JCYmsmLFCtq2bYulpeWjyqndoW5vGifYXCLbJWxFihThv//+Iy4uDqUUmzdvpnTp0rz11lv6rZPmz5//yrdLErmTt7c3kydPNnYYlC1bljFjxmS4fkhICBqNhuvXr7++oIQQ6UpKSSJwVSDJ2mTeq/oeDYo+bC1KTtYt4wHwxReQh7d8W7oUmjosAqBYu3aY2dpibm7O9u3bGTFixKOK8dfgxm4wsQDXhk85m8iIbJewVa9enbfffht/f3/KlSuHVqulf//+DB8+nI0bN1KiRAk2btzI8OHDjR2qUQQGBqLRaPQ/jo6OtGrViuPHj7+ya0RGRqLRaJ66PUZ6PvjgA/Lly8evv/762q/1Ks2bNw+NRkOJEiXSlK1duxaNRoOtra0RIhNCGMv4sPEcuHqAogWKMqHxYysSzJ+v29XAxwceToLLi+LjYcs/t6hlvwYA327d9GVly5alVKlSjypHBQMKXOqDqU3WBprLZLuEDeDLL7/k+PHjHDlyhN9++w0LCwsKFSrE5s2bOXXqFJs3b8bBIe+u49K4cWOioqKIiopiw4YNPHjwgHbt2hktnoSEBBYuXMjw4cOZPXu20eJ4UZaWlty+fZvQ0FCD40FBQRQpUsRIUQkhjOHA1QN8FfYVAEFtHusKTUiAsWN1z7/8Ms9ONADdYrnVTJdhbpKAW9265PfyIj4+nt69e7N//37DytId+spky4RNPJuFhQWurq64urri7+/Pxx9/zPHjx3nw4IG+zuXLl+nSpQsFCxakYMGCtGzZklOnTunLL168SJs2bXBwcMDa2ppSpUqxePFiAIo+3AevatWqaDQa6tev/8x4/vrrL7y9vRk1ahQREREcOXLEoFwpxXfffUeJEiWwsLDAw8ND32T+tGsFBgbS6uEg1lRjxoyhbNlHM4z27NlD06ZNcXR0JH/+/NSuXZudO3dm4k7q5MuXjx49ehAUFKQ/dv36ddasWUPPnj3T/bzlypXDwsICT09Pvv76a5RS+vKYmBjatGmDlZUVXl5eBudNdefOHfr374+zszN2dnbUq1fPaK2MQgidxJREAlfqukIHVx1Mfe/6jwpnzYILF6BsWejSxWgxZgfLlybR2EH374Vv9+4ArFu3jrlz5xoO49AmQ9R63XNJ2F6aJGw53N27d1myZAnlypXDysoKgLi4OBo0aIClpSWhoaHs3LmTwoUL07hxY+Li4gB49913iYuLY+vWrRw9epRp06ZRoEABAHbv3g1AcHAwUVFR/PXXX8+MYfbs2bzzzjtYW1vTvn37NK1sI0eOZNy4cYwYMYKjR4+ydOlSPD09X+haT372Hj16EBYWxu7du6lYsSJvvvnmC4376tOnD8uWLdOvzv3bb79Rs2ZNij8xA2zv3r107NiR9u3bc/jwYSZMmMA333zDTz/9pK8TGBjI6dOn2bRpEytXrmTBggVERkbqy5VStGzZksuXL7NmzRr2799P3bp1adiwoX5ijRAi640PG8/B6IMUK1jMsCv0/n34+mvd83HjdDsb5FEJCXBh4xYKmV3F0r0ohWvWBGDJkiU4OTkZzg69/h8k3gI7X7DzMVLEuYfsJfqE9FqTOnXqpE9w3nwz7f8SAgMDCQwM5Pr167z99ttpygcNGkTnzp25ePEiPXr0MCgLCQnJdIzBwcH6cVX379/H09OTtWvX6ssXL16MUoq5c+fqNyj+5ZdfcHZ2Zs2aNXTq1Inz58/ToUMHKlSoADxq6QL0s2sLFSqEq6vrM2M5e/YsYWFhLFy4EICePXvSqVMnvv32WywsLLh37x5Tp05l2rRp9O7dGwAfHx9q1KiR6Ws9yWBhRuDHH39k+fLlBAcH804mN2BO3Qpt8eLF9OvXj6CgIIYNG0ZycrJBvSlTplCvXj2+/PJLAHx9fTl16hTffvst77//PidPnmTdunVs376dWrVqAbpJMsWKFdOfY+vWrRw4cIBr167pk+xx48axevVqfvvtNz777LNMxS6EeHkHrh7g6zBdUhb0VhA25o+Nt/rpJ4iOhqpVIY9PeNu0CepY6ZbaKhvYDY2JCffv32f16tX07NkTU9PH0grpDn2l8u5/E3KwunXrcuDAAQ4cOMCuXbto2LAhTZs25eLFi4CuFejcuXPY2dlha2uLra0t9vb23Lp1izNnzgDw4Ycf8tVXX1GjRg0+//xz9u7d+0KxBAUF0ahRI32yVb9+faytrfU7URw7doyEhAQaNWr08h/8CTExMQwYMABfX1/s7e2xs7MjJiaGCxcuvND5+vTpQ1BQELt27eLixYt06NAhTZ2IiAh9Ipaqdu3aXL58mdjYWCIiIjAxMaFatWr6ci8vL9zc3PSv9+7dS1xcHE5OTvrvx9bWliNHjui/HyFE1klMSSRgZQDJ2mTer/Y+9bzrPSq8cwe+/Vb3/Kuv8vw6YuvnH6OU9T60ZrYUfZi8rlmzhri4ODp37mxYWRK2V0pa2J7wrBYva2vrZ5Y7Ojo+s9zT0/OFWtTSi8PH51HzcuXKlbG3t2fWrFmMGzcOrVZLxYoV9WPSHpc6WaNPnz40a9aMtWvXsmnTJmrWrMmIESMytfRESkoK8+bN48qVKwb/q9JqtcyePZvOnTsbjO3KDBMTkzTvTUpKMngdEBBAdHQ0U6dOxdvbGwsLCxo1akRiYuILXbNLly58/PHHDB8+nK5du+pbvx6nlNK3Wj5Jo9Fk6PNqtVpcXFwICwtLU5Y/f/7MBy6EeCm/7v2VQ9GHKF6wON80+sawcMYM3b6hdevC46v350HJyZCyZwlYg2uzdpjZ6FohExISqF69uuH+oXGX4PZB3cxQ57pGijh3kYQtF9BoNJiYmOjHp/n7+/PHH3/g6OioH5eWHg8PD/r370///v359ttv+f777xkzZgzm5uaALiF7luDgYG7cuEF4eLj+PQAXLlygVatWREZG4ufnh4WFBZs3b0536YynXcvJyYkDBw4YHHvy9fbt2/nhhx9o2VK3N110dPRLjQHLnz8/b7/9NgsWLGDSpEnp1vHz82P79u1p4vDw8MDOzo7SpUuj1WrZs2cPNR+O7bhw4QJXrlzR1/f39yc6OhoTExODrlIhhHGsOL4CgDH1xxh2hSoFCxbonn/2WZ5vXdsRlkw5880AVOnTXn+8Z8+eaSdoXVype3RtCvny7np1r5J0ieZACQkJXL16latXrxIREcH777/PvXv3aN26NQDdu3fHxcWFNm3aEBoayrlz59i2bRtDhgzRzxT98MMPCQ4O5uzZsxw4cIDg4GD8/PwAcHZ2xsrKivXr1xMdHc2dO3fSjWP27Nm0aNECf39/ypYtq/958803KVmyJEFBQdjZ2fHhhx8yYsQI5s6dy5kzZ9i9ezczZsx45rUaNmzI/v37CQoK4vTp00ycOJEdO3YYXN/X15fff/+dY8eOsWfPHrp06WKQOL6IX375hevXr1OlSpV0y4cMGUJoaChjxozh5MmTLFy4kO+++04/7qxkyZI0b96cAQMGsHPnTg4cOEBgYKBBa13jxo2pVasWbdq0Yd26dZw7d46dO3cyevTodFvdhBCvz92Eu2w7vw0TjQlvlnii6+7gQYiIAEdHaNrUOAFmI9sW7iO/6S3irb2wf/gf8Li4uPT/c39JlwTjabwlp3IbSdhyoE2bNlG4cGEKFy5M9erV2bNnD0uXLtVPmLC2tmbbtm0UK1aMjh07UqpUKQICArh16xYFCxYEdN1y77//Pn5+fjRp0gQXFxf9ThKmpqb88MMPzJ49Gzc3t3R3lYiOjmbNmjXpTrIA6NixI3PnzkWr1fLNN98wbNgwxo0bR+nSpenQoQOXLl165rWaNWvG6NGjGTVqFJUrVyYyMpJ3333X4BpBQUHcu3ePypUr06VLF3r37o23t/dL3VtLS0sKFSr01HJ/f3+WLl3K8uXLKVu2LMOHD2f48OEMHjxYX2fevHkULVqUhg0b0rp1a7p162YQl0ajYe3atTRs2JB+/fpRsmRJOnXqxIkTJwzGugkhXr+NZzeSpE2ihkcNHKyeWN9zkW4lfzp1ytPrrqW6tXMTAA41G+uHhkyZMgVXV1d9Dw8ACTcgJhQ0puDeKr1TiRegUS86yCgHqFKlylPXtoqIiKB06dJZHJEQIi+T3zvZT9+/+zJn/xy+bvg1I+uMfFSg1YKXF1y6BNu3wxOTjfKa85FaVjdtjINZNA1/X4xrpXIA1KtXj3v37hlOXDs7H/4LBNfG0HCjcQLOoZ6Vt0gLmxBCiDxJKcXaU7qZjC1LtDQsDAvTJWve3vBwPGpetn7eERzMookzdcWlom4B83v37rFz504aN25sWDm1O9RDukNfJUnYhBBC5En7r+4n6l4UHvk9KO9S3rDw4dqSdO2a5ycbAFzYqOsONS/3qDt027ZtJCUl0eTx2bPJ9x/tbuCRt9ese9UkYRNCCJEn/XPyHwDe9HnTcLmehARYtkz3/OHWS3nZgweKQtd0XZtVuj9qTdu4cSOWlpbUrl37UeWoDZASD4Wqg7V7Voeaq8myHkIIIfKktacfdof6PtEdGhysW3utfHkoU8YIkWUvWxafwsXsAnE4UKqpv/7422+/TcmSJbG0tHxU+aLMDn1dJGETQgiR51y7f41dl3Zhns+chkUNt7nTzw7t1i3rA8uGjizfhCeQ4NUAk3z59Mdr1apluPOLNgkur9Y9l/Frr1ye7hLVarXGDkEIkUfI75vsZcXxFSgU9b3rY2tu+6jg1i34+2/d865djRNcNqIUmJ3RdYeWavtorNqBAwcICwsz/HN9bTsk3Yb8pSG/bxZHmvvl2YTNxsaGy5cvk5iY+MLbJwkhxPMopUhMTOTy5cvY2Ng8/w3itYtLimNs6FgAepZ/YoX+6dMhPl63UG6RIkaILnvZv/USriYnidfaUKdHdf3xKVOm0K7dE61o1x4ubl5YFhl+HfJsl6iHhwfXr1/n/PnzJCcnGzscIUQuZmpqir29PY6OjsYORQDT/pvG5buXqeRaia7lHmtFi4uD77/XPR82zDjBZTPbg7bgCNxxqou5lW4nGaUUGzdupEmTJpiYPNbuc2OP7rFQtawPNA/IswmbiYkJzs7OODs7GzsUIYQQWSTmfgwTtk8AYHLTyZhoHks4goLg+nWoWhUaNDBShNnLg4NbAPBq+mic35EjR7h69SpNH9+uSym4sVv3XBK21yLPJmxCCCHyni9DvuRu4l3eLPGm4WSDpCSYPFn3fPhwWXsNOHvkFu4pe0nGlMYD6uiPb9iwAcBw/bUHlyH+KpgXBNviWR1qnpBnx7AJIYTIW05cP8Eve3/BRGPCxMYTDQv//BPOnwdfX0hn/+S8aPMvIZhotNywrUZ+Jzv98dDQUPz8/PDw8HhUObV1zaGqJLuvibSwCSGEyBOGbRpGikqhn38/yjg/tr6aUjBB103KZ5/BY0tX5GXX/92KDeBUq5HB8WXLlnHp0iXDyvrxa1WzJrg8SFrYhBBC5Hrbzm9j1YlV2JjZ8GX9Lw0L166FI0fAzQ3eecc4AWYz16Me4Bqnm/XZcKDheD5zc3OKFStm+AYZv/baScImhBAiV9MqLUM3DAXg05qfUtiusGGFb7/VPX78MVhYZHF02dP6mTuxMIknxqwcbiVd9Md/+OEHPv/8c8PKSgs3w3XPpYXttZEuUSGEELnan0f/ZM+VPbjaujKk5hDDwh07ICwMChSA/v2NEl92dGHTFrwA20qGu0AEBQXh4OBgWDn2JCTFgrUHWD2RDItXRlrYhBBC5FrJ2mRGbh4JwLgG4wx3NQCYMkX3+O67kD9/FkeXPT24n4LDjRAAavd+lLBFR0dz8OBBw+U8AK4/XDDXQVrXXidJ2IQQQuRay48t59ztc/gW8iWwYqBhYWKibqN3gEGDsjy27GrLHxHY5bvFbdwpWfvREh2hoaEANGz4WKubUnByuu65W/OsDDPPkYRNCCFErqSUYsp/uha0j9/4GFOTJ0YB7dyp292gTBl4fImKPO7w3/8BoLzeQPPYEh2hoaHY2tri7+//qPLVjXBrP1i6QNGeT55KvEKSsAkhhMiVdl7aye7Lu3GwcqBnhXSSiY26Tc15fAHYPE4pSDy5CwDfZm8YlFlZWfHWW29havpY4nvs4XIoJT+CfJZZFGXeJJMOhBBC5EpTdupa1wZVGYS1mXXaCps26R4bN87CqLK3Y4cT8dTsA6Bm1+oGZZNTd4JIdX0XRG8Fs/xQQrqUXzdpYRNCCJHrnL11lhXHV2BmYsZ7Vd9LW+HWLdizB8zMoF69rA8wm9o8/wAWJvHctfLFxrmQ/nhycnLaysceLodSYhCY22dRhHmXJGxCCCFynR92/YBWaelWrlvaddcAQkJAq4UaNcDWNm15HnVhm278Wv7yhq1rw4YNo1y5cmi1Wt2BO8fh0kowsYCSH2ZxlHmTJGxCCCFyldvxt5mzfw6gm2yQrtTxa9Idqnf7Ntjd0I1f8+9gOH5t27ZtODg4YGLyMG2ImAQoKBYoa69lEUnYhBBC5Cqz983mXuI9GhZtSAXXCulXkgkHaaxffY9ilofRko9i9aroj9+9e5d9+/ZRt25d3YG4SxD5G2hMoPRQI0Wb90jCJoQQItdI1ibzw64fAPjkjU/SrxQZCadPg709VKmSfp08aM+ycPJpUkhyKofZY93E//77L1qtlnqpY/2OTwVtEnh2BDsfI0Wb90jCJoQQItdYcmQJF2MvUrJQSVqUaJF+pdTZoQ0agKkslgC64Xyxh3Tj1zzrGI5fCw0NxdTUlBo1akDibTj9i67Ab1gWR5m3yZ9UIYQQucL9xPuM2DwC0G3ybqJ5SpvE2rW6xye3WMrDDh8Gdw4BUKaF4RZTdevWxcrKChsbGzg7D5Lvg3N9cKiU9YHmYZKwCSGEyBW+3fEtF2MvUsm1UtptqFIlJMCGDbrnLVtmWWzZXciWJLwsjwNQqIyfQVnz5s1p3vzhtlPn/9Q9enfNyvAE0iUqhBAiFzh36xwTd0wE4McWP5LPJF/6FUND4f59KF8eihTJwgiztwMbz2BukoC2gCfm9o/WVLtw4QLHjh1DKQWJt3RbUWnygUc7I0abN0nCJoQQIsf7ZMMnJKQk8E75d6hVpNbTK/7zj+6xVausCSwH0Goh5tBRABzLlTUomzFjBhUqVOD+/ftwcSWoZHBpAJZORog0b5OETQghRI624cwGVh5fiY2ZDd82/vbpFZWC1at1zyVh0zt6FFxSdAlbkWplDMpCQkKoVq0atra2cOFhd2iRTlkdokASNiGEEDlYYkoiHwbrVtr/X93/4Wbn9vTKx4/DuXPg6AjVqmVRhNlfSAgUtToCQKGyjxK2u3fvsmfPHho0aAAJN+HqJukONSJJ2IQQQuRYP+3+iePXj1PCoQQfvfHRsyuvWaN7fPNNyPeUMW550LYtiXhZngDAwe/RhIPt27eTkpJC/fr1ddtQqWRwaQiWjsYJNI+ThE0IIUSOdPXeVcaEjAFgWvNpWJhaPPsNqQmbzA7V02rh9H+nMNUkY+le1GDB3JCQEMzMzKhZs6Z0h2YDkrAJIYTIkUZsHsHdxLu0LNGSN0u8+ezKN2/Cjh26hXJl/TW9Y8fAIV43fs2louH4tWHDhrFx40as8z2Aq5t13aGe0h1qLJKwCSGEyHHCr4Qz78A8zPOZM7XZ1Oe/Ye1aSEmBOnWgQIHXHl9OERICxdIZvwbg4OCg244qtTvUtTFYFMr6IAUgCZsQQogcKHXNtfervU+JQiWeXTkpCcaN0z3v2PE1R5azLF2ciL/dVgCcK1fWH//vv/+YMGECd+/ehaj1uoMy2cCoJGETQgiRo5y/fZ7lEcsxNTHl4zc+fv4bpk+HkyfB1xf69Hn9AeYQJ09C/KGt2JvexK64LwUfm3CwePFixo4di5mpKUSH6A66NDROoAKQhE0IIUQOM33PdLRKS6cynXDP7/7sytevw5df6p5/9x2Ym7/+AHOIoCCoX3A5AL6dOqDRaPRlmzdvpnbt2lgmnIWEa2DlBnY+xgpVIAmbEEKIHORe4j1+3fcrAB9W//D5b/jiC7h9WzfRQGaH6iUnw9+/XaGczb9gao73YwsJX716lSNHjtC4cWOICdEddK4PjyV0IutJwiaEECLHWHBwAbfjb1PTsybV3J+z+O3hw/DLL7o116ZMkYTjMcHBUDppBSYahVfTxlg8NhFjy5YtADRq1Oix7tAGWR+kMCAJmxBCiBxBq7R8v+t7AD6q/tGzKysFH3+sW2hs4EAoU+bZ9fOYoDkp1CvwFwDFO3QwKLt48SKFCxemYoXyj1rYXOpnbYAiDUnYhBBC5AjBp4M5eeMknvk9aVf6OTMWV6+GzZuhYMFHY9gEANHRELnlXxzNrmLl5onLE9t0DRs2jAsXLpDv3glIuA5W7mBb3EjRilSSsAkhhMgRpv03DdAt5WFqYvr0igkJMGSI7vmYMVBI1g573O+/Q938y4CHkw1M0qYCpgazQ+tLd3I2IAmbEEKIbO9ozFE2nt2ItZk1ff37Prvyjz/C6dNQqhQMGpQ1AeYQSsGSoOv424WAST6KtmljUD5nzhxq1arF7du3DSccCKOThE0IIUS2lzp2rVfFXhS0Kvj0itevP1okd8oUMDPLguhyjt27wTn6b0w1ybjVrYu1s7NBeXBwMBcvXsQ+vx3EhOoOyoSDbEESNiGEENlabEIsCw8vBHTdoc+0aBHExkLjxtCiRRZEl7MsXQq17NcA4NOhvUGZVqtl69atNGzYEE1MiG78mo032BbL+kBFGpKwCSGEyNaWHFlCXFIcdb3qUtKx5LMrL12qe+zV6/UHlgPtXHMOL8sTaKzsKFy7tkHZ4cOHuXHjBg0bNoQzQbqDRQNk/Fo2IQmbEEKIbG3O/jkA9K30nLFrly/D9u1gYQGtW2dBZDnLqVPgfEO3L6hXk4bke2LXh61bdXuKNqhVCS4uBzRQXBLf7EISNiGEENnW0Zij7Lq8i/wW+eng1+HZlZfrtlnizTfBzu71B5fD/PMPvJE/GACvFs3TlBctWpQ+ffrgmbINtAng2hhsvLI6TPEUz5gXLYQQQhhXautat7LdsDazfnblP//UPXbs+Jqjypm2rzhNe8tTKMv8uL7xRpryNm3a0KZNG1jnrztQrHcWRyieRRI2IYQQ2VJCcgILDi4AoI9/n2dXvnwZduwAS0t4bF9MoRMbC5rjG8ARPBo2TtMdeuPGDbRaLU75LsGt/WBeEDzbGidYkS7pEhVCCJEt/X3ib248uEF5l/JULlz52ZWX6RaCpUUL6Q5Nx4YNUNVW1x1aok2zNOVz5szBxcWF63t/0h3wfgfyWWZliOI5JGETQgiRLT0+2UDzvJmKqbNDO3V6zVHlTCF/nsLD8gwpFgVwrV49TfnWrVspVaokjrd1+4tS/DktmiLLScImhBAi27lw5wIbzmzAIp8F3ct3f3blS5ekO/QZUlLg7m5d65pT7caYPLGYcFJSEmFhYTSo7AlJt8GhMhSsYIRIxbNIwiaEECLbmbt/LgpFu9LtcLByeHblx2eH2tq+/uBymN27FeVMdct5VOiSdnZoeHg49+/fp0GxKN0BaV3LliRhE0IIka2kaFMIOqBbuLVPpQwkD6kJm8wOTdf2NZdwszhHkqk9LtWqpilfsWIFpqam1Hc7Cpp84NXVCFGK55FZokIIIbKV9WfWc+HOBYoVLEbDog2fXTk6WrdYrrk5tGyZNQHmMFG79uEO5POujIlp2n/2P/nkE94oXQBHi1FgXwbMC2R5jOL5pIVNCCFEtjJr7ywA+vv3x0TznH+m/v4blIImTWR26FOoi/sAcH/DP91yV1dX2td8eO8cqmRVWCKTJGETQgiRbVyOvcyak2swNTElsGLg89/w18NZje3avda4cqobN8AtRZew+TVLm7CNGjWKv//+G27u1R1weM7yKcJoJGETQgiRbQTtDyJFpdCuVDtcbF2eXfnOHdi8GUxM4K23sibAHCZ82y3cLc6SjAWOZUsblF28eJEJEyawZ88eSdhyAEnYhBBCZAsp2hRm758NQP/K/Z//hn/+gaQkqFMHnJxec3Q50/FN+wG4X6B8mt0N5s2bh1arpVePLhB7TDfhoEB5Y4QpMkASNiGEENlC6mSD4gWLP3+yAcCKFbpH6Q59qluHdN2htqUqGRzXarXMmTOHRo0aUazgHVBa3YQDUytjhCkyQBI2IYQQ2ULqZIN+/v2eP9ngwQNYu1b3vG3b1xtYDmZxTdfC5tPAcPza5s2bOX/+PH379n2sO1QmHGRnkrAJIYQwutTJBmYmZvSq1Ov5b9i4EeLioHJl8PJ6/QHmQDFX4nE3OYJWmVCxlWELW3JyMnXq1KFt27Yyfi2HkIRNCCGE0aVONmhbqi3ONs7Pf4PMDn2uPasOY6pJ5kY+X6wKGO4A0aJFC7Zt24alpaUkbDmEJGxCCCGM6l7iPX7a8xMAAyoPeP4bkpNh9Wrdc0nYnuri7qMAJLsa7gt6+vRp7t27p3uhTYbYCEAjEw6yOUnYhBBCGNVPu38i5n4Mb3i8kbHJBjt3ws2bULw4lC79/Pp5VNzlCwDYFClmcLx37940aNDgYaVLoFLAyk0mHGRzkrAJIYQwmjvxd5i4YyIA4xqMQ6PRPP9Nqa1rb70FGamfR6kbuoTNsUQR/bFLly6xfft23kpdt+5+pO7R1jtrgxOZJgmbEEIIo5n631Ruxd+inlc9GhVtlLE3pSZsrVu/vsByAat4XcLmWeFRwrZ06VKUUnTu3Fl3IDVhs/HO2uBEpsnm70IIIYziRtwNpuycAmSide3UKTh+HAoUgNq1X2+AOVhyQiL5VRQp5KNEFTf98cWLF+Pv74+vr6/uwP3zukcbmWmb3UkLmxBCCKOY/O9k7ibepVnxZtTxqpOxN6W2rrVoAWZmry+4HO7S4UuYaLTcTHHDwUm3w8GFCxfYvXv3o9Y1kBa2HERa2IQQQmS56HvR/LD7B0DXupZh0h2aIef26bpD75l56o95enpy8OBBXF1dH1W8F6l7lIQt25OETQghRJabsH0CcUlxtCnZhqruVTP2plu3ICwMTE2hefPXG2AOdzVCl7Cl2D8av6bRaChf/omlO/QtbNIlmt1Jl6gQQogsdSn2EjPCZwAwtsHYjL8xOBhSUnSbvRcs+Jqiyx1iz+sSNgtXXcJ24sQJAgICOHfu3KNK2mTdsh4ANkWePIXIZiRhE0IIkaW+3vY1CSkJdC7TmfIumVis9e+/dY/SHfpcSTG6hM3eW5eILV68mN9++w0LC4tHlR5cAZUMVoUhn6UxwhSZIAmbEEKILLPr0i5m7ZuFicaEMfXHZPyNcXHwzz+655KwPZd5rK4lzdVP19W5evVqatWqhZvboxmj3D6ke7Qt9uTbRTYkCZsQQogs8SDpAQErA9AqLZ+88QmlHEtl/M1//QV370L16uDj8/qCzAUSbt/GNuUKCVpLvP29iI+P59ChQ9R+chmUyw8ncLg2yfogRaZJwiaEECJLfL7lc07cOEFpx9KMa5iJmaEAc+fqHnv1evWB5TK3jh8H4EJ8STyL5OPQoUMkJSVRpUqVR5WUFi497GL2aGOEKEVmScImhBDitQs7H8bU/6aST5OP+W3nY2maiTFTkZGwZQtYWkKXLq8txtzi6oEIAC4ll8LBAWJjYyldurRhwnYjHOKvgnURKFDhKWcS2Yks6yGEEOK1up94n16reqFQDK89POPLeKSaP1/32L492Nu/+gBzmaj9uoTtrnVpNBpo3Lgxx44dM6x0eZXu0UP2Y80ppIVNCCHEazVs0zDO3DpDOedy/K/u/zL3Zq0W5s3TPZfu0AyJPalL2FKc/Z5eKbU71P2tLIhIvAqSsAkhhHhttpzbwvQ90zE1MWVBuwVYmFo8/02PCw3VdYl6ekLDhq8lxtwkOS6O5JhzJCtTbLxLEBcXR+HChZmf2koJcO8s3DkCZvnBuZ7xghWZki0Tttu3b/P2229TqlQpSpcuzc6dO7l58yZNmjShRIkSNGnShFu3bhk7TCGEEM8QmxBL71W9Afhf3f9R0bVi5k+S2roWEAAm2fKfrGzl1okTaFBcTiiGexFzDhw4wNWrVyn4+ELDqa1rhVtAPnPjBCoyLVv+6f/www9p3rw5x48f5+DBg5QuXZoJEybQqFEjTp06RaNGjZgwYYKxwxRCCPEMQzcM5fyd81QuXJkRtUdk/gR378KyZbrngYGvNLbc6laErjs0Mt4PT08IDw8HMJxwcCl1/JrMDs1Jsl3CFhsby7Zt2+jTpw8A5ubmFChQgFWrVhEQEABAQEAAK1euNGKUQgghniX4dDC/7vsV83zmzG87H7N8Zpk/yZ9/6hbMrVsXihd/9UHmQjcfJmzn40vj4aFL2AoXLvxowdyEm3AtDDSm4NbCiJGKzMp2CdvZs2dxcnKiV69eVKpUib59+3L//n2io6MpXLgwAIULFyYmJsbIkQohhEjP3YS79F/dH4Cx9cdSxrlM5k+SkACTJumey2SDDFFKEfOwRe3cg9J4esL+/fvx9/d/VOnCUlApurFr5gWME6h4IdkuYUtOTmbfvn0MGjSI/fv3Y2Njk6nuz1mzZlGlShWqVKnCtWvXXmOkQggh0vO/rf/jYuxFKheuzJCaQ17sJJMmwYkTULIkdO36agPMpa7t28e9Cxe4leREjFkFypSBVq1a0a5dO10FpYUT03TPi/cxWpzixWS7hM3DwwMPDw+qV68OwNtvv82+fftwcXEhKioKgKioKJydndN9f//+/QkPDyc8PBwnJ6csi1sIIQTsvrybH3b9QD5NPn5t/SumJi+w3Ofp0/DVV7rnM2aARSZnluZRZ//6C4Btd9pSr4EppqbwzTff6IcYEbUeYo+DtQcUeduIkYoXke0SNldXVzw9PTlx4gQAmzdvxs/Pj7feeks/LXn+/Pm0aSODJYUQIjtJSkmi3+p+KBSf1PiESoUrZf4kSsF77+m6RHv0gAYNXn2guVDS/fucX78egNDb7WjcGBISEkhISHhU6fhU3aPvYDB5gTGFwqiy5U4HP/74I927dycxMZFixYoxd+5ctFotnTp1Ys6cORQpUoSlS5caO0whhBCPmbJzCoeiD1G0QFFG1xv9YidZsgQ2bICCBWHy5FcbYC52ITiYlAcPOJNUmehELxo3hjVr1vD2229z6NAhynlq4OpGyGcNPv2NHa54AdkyYatYsaJ+KvLjNm/ebIRohBBCPM+Zm2cYEzoGgBktZ2BjbpP5k9y+DR9/rHs+cSI8ZeiLSOvMw+7QjTHtcXfXDf37559IADw9PeHEUF3FYoFgXjD9k4hsLdt1iQohhMhZlFIM/Gcg8cnxdC/XnWY+zV7sRKNGwdWrUKsW9O79aoPMxe6cOcP1AwfQmlmzO7YpjRrptgeNjIzE3t6eApaJcO53XeWSHxo3WPHCJGETQgjxUn4/9Dubzm7CwcqBqc2mvthJdu/WTTAwNdU9yq4GGXZ2xQrdo0ULEpQ1jRvrjkdGRuLt7Q2nZoA2AdxaQX5f4wUqXor8jRBCCPHCrsdd5+P1um7MKU2n4GTzArPzk5NhwADdhINPPoFy5V5xlLmXNimJc3/rtppaeqoDAI0a6coiIyPx9vKEUz/rDpT62BghilckW45hE0IIkTMM2TCEGw9u0LBoQ3pW6PliJ/npJzhwALy84IsvXml8ud2VsDDib9zA1LUYR46Vx88PUjc1GDhwIK4cg/g1UKA8uMiM25xMEjYhhBAv5OSNkyw4uABLU0t+afULGo0m8ydJSYGxY3XPf/oJbF5gskIeFvnPPwBccGgPaPTdoQDvvfcebKwD19C1rr3I9yOyDUnYhBBCvJC1p9YC0L50e3wcfF7sJBERcOuWrnWtVatXGF3up5QiZs8eAH77T9d61rGjriw2Npbr0ZcpErMLUw2y0XsuIGPYhBBCvJD1Z3QLtTYv3vzFT7Jzp+7xjTdeQUR5S+y5c8TfuIHGrhCHLnnh46ObYAsQHBxMcV8/jl1MgvylZSmPXEASNiGEEJn2IOkBIZEhADQt3vTFT/Tff7pHSdgy7drD9UovaaoCGgIDH/V6nj17FoCiToBjDaPEJ14tSdiEEEJkWtiFMOKT46nkWgkXW5cXP1FqwlZDkorMin6YsG05UxmNBno+Nufj7NmzOBWwwM4KcJRkODd4qYTt9OnT7Nixgzt37ryqeIQQQuQAwaeDAWhW/AUXyQXdzgbHjoG5OVSs+EriyiuUUsQ8TNiOxFalcWPw9HxUfvbsWYo5K90LaWHLFV4oYVuzZg3FixenZMmS1K1bl7179wIQExODj48Py5Yte6VBCiGEyF7049d8XmL82u7dukd/f7CweAVR5R33L13iQXQ0D7DnckJxevUyLD939jTFCiWCqZ1uDJvI8TKdsIWEhNCuXTscHBwYPXo0Sil9mbOzM8WLF2fx4sWvNEghhBDZx8U7Fzl27Ri25rbU8HyJ1hsZv/bCHrWuVSG/vQlt2xqWTxzWkQGNAMfqYJIvy+MTr16mE7axY8dSoUIFdu3apVvj5Qk1atRg3759ryQ4IYQQ2U9q61rDog0xz2f+4ieS8WsvLOZhz9bxuMp06QJWVoblHappqFcaKCTJcG6R6YQtPDyc7t27Y/KUfd48PDy4evXqSwcmhBAie3oly3lotdLC9hKi9+ha2I7HVUnTHXrlyhVCtwTzIBEZv5aLZDphS0lJweIZYw2uX7+OuflL/I9LCCFEtqWUYvPZzQA083mJCQfHj+sWzC1c2HC0vHiuexcvcv/SRR6k2GDlXYpq1QzL/165kvpDDxNzB12XqMgVMp2wlS5dmrCwsKeWr1mzhgoVKrxUUEIIIbKni7EXuRV/CydrJ4oVLPbiJ9qwQffYoIFsmZRJZ/76C4C9dxvSpWu+NLcvLGQ9LvZQpIg7WBQyQoTidch0wtanTx+WLVvGnDlz0Gq1AGg0GuLi4vjggw/YuXMn/fv3f+WBCiGEML7D0YcBKOdS7uVOFKxbFoQWLV4yorxFm5zM2ZUrAdh6+23atTMsV0qxJTSMhn6gsZfZoblJpvcSHTRoEDt27KBfv34MGTIEjUZD165duXHjBikpKfTq1Yvu3bu/jliFEEIY2eEYXcJW1qnsi58kLg5CQnTPm77ELgl50JWwMB7ExHAlwZukwpUp+8TXcPz4ca7G3KLhW8hyHrnMC23+/vvvv9OhQwd+//13jh8/jlKK6tWr07NnTzp06PCqYxRCCJFNpCZsL9XCFhoKCQlQuTI4O7+iyPKGMw/XOQ253YG2gZo03aFbt24FoGEZIH+pLI5OvE4vlLABtGvXjnZPtsUKIYTI1fRdos4vkbBJd+gLiYuO5sq2bSQrU8Jut2FEOv8E9+vXj8rJMyla6DBIl2iuInuJCiGEyJCklCSOXz8OQBnnMi9+onXrdI/NX2JZkDzo7IoVKK2WvXcbYuVYKN3l68xMTanudknX8iYtbLnKc1vYFixY8EIn7vn4LrRCCCFyvJM3TpKkTaJogaLYmtu+2EnOnIFTp8DeHqrLkhMZpbRa/ezQrbfepk13eHI51IiICH6d8T2flLiFh2t+sHQ1QqTidXluwhYYGIhGozHYgup5NBqNJGxCCJHLvJLxa+t1i+7SpAmYvvConDzn6n//cf/yZe7gxpH7Nfi2bdo6a9euZeqPvzDkR3QTDmS5lFzluX9bUgcwCiGEyNtk/JrxpE422BDTHls7Exo2TFtny5YtlCzmgrtDtHSH5kLPTdjq1auXFXEIIYTIxpRShJ4PBV4iYbtxAzbrdkmg2UvskpDHxN+8yaXNm1EaE7bdbkfLt+HJDYfi4+PZtm0bPZp5A5Kw5UYy6UAIIcRz/XHkD3Zc3IGDlQNNijd5sZOMHatbg61JE3B3f7UB5mLn/v4bbXIyJ1PqcDPZla5d09ZZvnw59+7do0P5a7oDznWyNkjx2j23hW3btm0A1K1b1+D186TWF0IIkbPdenCLj9d/DMDExhNxsHLI/ElOnoSff9aNq5o06RVHmHsppfTdoWsuvY27O7z5Ztp6N2/epHKF0jQoFqGbbCCbvuc6z03Y6tevj0aj4cGDB5ibm+tfP41SCo1GQ0pKyisNVAghhHEM3zScmPsx1ClSh16Ver3YSYYNg+Rk6N0bZL/pDLu2bx+x584Rb+rEgXt1+d+Q9OdqvP/++wyuE4XmWAR4tgONdKDlNs9N2IKCgtBoNJiZmRm8FkIIkfvtuLCDWftmYWZixsxWMzF5kUQgNBRWrgRraxg37pXHmJultq5tjGkHJqb07Zu2zvnz5yni6Ynmkm7ZDzxlx6HcKEPLejzrtRBCiNwpKSWJgf8MBGBYrWH4Ofll/iRaLQwZonv+2Wfg5vYKI8zdEmNjubBhAwBbbrSjZUvw8DCsEx8fj7+/Pz06vcm0OifAohA4y2TB3CjT/1UaO3YsR44ceWr50aNHGTt27EsFJYQQwvi+2/kdR2KO8P/27js8iqoL4PBv00hPCAklEHpL6L33LoI0AVFAEBBBVFSKIip+FkBFUHoRqYpIE0Eh9N47hNAhCSG9992d749JAoEEUnZTz/s8+2Szc2fmbBbI4ZZzqxSvwqdtPs3eRdavh7Nn1UTt448NG2Ahd2/HDnTx8dzWNScwqTxvv/1sm82bNxMaGsrLDU3VF8q+AiZS364wynLC9uWXX3Lp0qUMj1+5coUZM2bkKCghhBB5607YHWYcVP8tX9RzEVbmVlm/SFwcfJqc6H3zDdjYGDDCwu3JxQY7/ftTvnz6O3ktXbqUypUr07HMefUFGQ4ttAw+KzE+Ph4zqV4thBAFlqIojNsxjnhtPK/XeT37ZTx++gl8fKB+fZDdb7Ik7No1wq5fJ8HUgTNRnRk1CkxN07bx9vbm4MGDjBneH5PIS2BuD6U75U3AwugylVlFRkYSHh6e+n1ISAgPHjx4pl1oaCjr1q3Dzc3NYAEKIYTIXX9e/ZNdt3fhaOnIj11/zN5FAgLgu+/U5z/++OzGl+K5biX3rh0I7o1iYsFbbz3bZtWqVZiZmfFm+2LgC7i+DKbFnm0oCoVMJWw//fRT6rw0jUbDBx98wAcffJBuW0VRmD17tsECFEIIkXsSdYlpaq6Vsi2VvQv9738QHQ0vv0y6+yiJDGnj47m3YwcAe0MH0KtX+ms1vvjiC3r06EGpmORFHeVlOLQwy1TC1r59e0BNxr766iv69u1L3bp107TRaDTY2trSvHlzWrZsafBAhRBCGN+uW7vwj/bHw8WDtxqm062TGQEBsHy5+nzmTMMFV0Q8On4cbUwMgSYe+CVU5ZcMRpOLFStGm7rOsOM0mNlBmXQmuYlCI1MJW7t27VL3FL1//z5jx46lWbNmRg1MCCFE7lt3eR0AQ+sOzV7NNYC5cyEhAV55BWrVMlxwRYTf/v0AHArogIWFupPX086dO8fGjRuZ2CWakgAVBoGZda7GKXJXllcHrFy50hhxCCGEyGNRCVH87f03AK/VTmfDysyIiFC3oAKYOtVAkRUdil6P34EDAJyN7Ei7dmBr+2y7Q4cOMXPmTD6sUQosgMpv5maYIg9keznnjRs3uHXrFiEhISiK8szxYbIiSAghCpQt17cQp42jTfk2VHCskL2LLF4MkZHQvj00b27Q+IqC4EuXiA8JIdbclQcJNfiwZ/rtbty4gaODLc7mAWBbFZxlKlJhl+WELSAggOHDh+Pp6QmQbrKm0WgkYRNCiAImZTj09TqvZ+8C8fFqKQ+Q3rVs8tu3D4CTYR0BDT0zSNhu3rxJdVcLNBrU3jXZMrLQy3LC9u677+Lp6ck777xDx44dKVGihDHiEkIIkYseRT9iz509mJuY82qtV7N3kd9+UxccNGgAXbsaNL6iwjd5/tqx4A7UqAFVq6bf7saN67QtHw5ooNLQXItP5J0sJ2yenp6MHTuW+fPnGyMeIYQQeWDDlQ3oFT0vV38ZJyunrF9Aq4Xvv1efT50qPT7ZEHnvHpF37qA1t8c7thHvZ9C7lpiYSGR4CNWb6qF0Z7Apn7uBijyR5YRNr9dTr149Y8QihBAij6QMhw6pPSR7F/jrL7hzB6pUgf5SDyw7fJOHQ68ntUWHeYbDoRYWFoRuqIM28BRUejP3AhR5Kstrttu0acPFixeNEYsQQog8cDPkJqcfnsbWwpZeNXpl/QKK8rje2uTJz+6hJDIlZf7aXp+O2NlB69YZNIzwQhN6CnNLO3Drm3sBijyV5YRtzpw5bNmyhU2bNhkjHiGEELlsxfkVAPRz74e1eTZqeXl6wsWLULq07BmaTfFhYQRduIBiYs6lmNZ07QoWFum3XTXvY4YuBF25V6X2WhGS5SHRd955B1tbWwYOHIirqyuVK1fG9Kn/TWk0Gvbu3WuwIIUQQhiHV5AXP51QV3a+3ejt7F3k11/Vr+PHg6WlgSIrWkKvXgVFIahYXeL1NvTokUFDvZYl6z2JjgXTaiNzNUaRt7KcsN25cweNRkP58uokx/Q2gRdCCJH/6RU9o7ePJlGXyOiGo2nplo1aXhERsG2b+nyorFbMroibNwG4FlQdSH93A4A7J37juHcSM4c7S+21IibLCdu9e/eMEIYQQojctuTMEo76HKW0bWlmd5mdvYts2qTWX2vXDipks9iuIPzGDQBuR1WnenUon8HCz3W/zgHgtTdGyUrcIiabG8UJIYQoyHwjfZmyZwoA83vMx9HSMXsXWrNG/Sq9azkSntzD5pNQLcPeNSUukHU7vWhbE8q3ei8XoxP5Qba3prp79y579+4lICCA119/nYoVK5KYmMijR48oXbo0FhnNlhRCCJGnFEVh/M7xRCVG0admH/q598vehR48gAMHoFgxGDDAoDEWJXqtlojbtwHwTaiaYcKWcGMVXetAi6b1wapM7gUo8oVsJWxTpkxhzpw56HQ6NBoNLVq0oGLFisTHx+Ph4cHXX3/NBx98YOBQhRBCGMImr0387f039sXsmd9jPprsDq2tX69+7d0bHBwMF2ARE/XgAfrERIKTypCosaN9+3QaKQqWfmv4eTjQZnouRyjygywPiS5ZsoTvv/+e8ePHs3v37jR7idrb29O7d2+2b99u0CCFEEIYRlhcGO/ufBeAWZ1nUda+bPYupCgyHGogKQsOfOKr07Rp+rmvLugUR05eRm9eAlxfzuUIRX6Q5YRt4cKF9O3bl7lz59KgQYNnjtetWxdvb2+DBCeEEMKwJnlOIiAmgNblWzOm0ZjsX+j8ebh2DZydoXt3wwVYBKUsOHje/LV9f35Dm6/g7wctwFSmHBVFWU7Ybty4QZeM/kQBLi4uBAcH5ygoIYQQhrf/7n5WnF+BhakFy3otw0STg3VnKb1rgweDublhAiyiwm/dAsAnvhqdO6fTQBfPur/+w94Kur/+Re4GJ/KNLP9ttbS0JCYmJsPj9+/fx9HRMScxCSGEMDCdXsfYHWMBmN52OjWda2b/YklJ8Pvv6nMZDs0RRVEIvqqOSgWbVqd582fbxN3ayOaTSfRvXQLL0o1zOUKRX2Q5YWvatClbtmxJ91h8fDxr1qyhVatWOQ5MCCGE4VwLusaNkBu42rkyudXknF1s0yYICAB3d2jSxDABFlEhly4R7+9DlNaBWm0rpdtZ6bn1V6LiYfAgWYlblGU5YZs0aRLHjx9n6NChXLp0CYBHjx6xa9cu2rdvj6+vLx9//LHBAxVCCJF9R32OAtC+YnsscjoHat489et770nx1hzyXrcOgAPh/XljWDrZmqLw7/5T2FtB+17Z3DpMFApZLuvRuXNnFi1axPvvv8/65CXdQ5O7xC0sLFi2bBktWrQwbJRCCCFy5JjPMQBalsvhdkYnT8KJE1C8uAyH5lBcUBAP/tuFXjHhlP41fuuZTqNIb+YNieWdrk5YuNTP7RBFPpKtOmxjxoyhd+/ebNy4kevXr6MoCtWqVWPgwIGULZvNJeJCCCGMJqWHLVv7hT4ppXdt9GiwsclhVEXbrY0bUXRazkZ1ousAV4oVS6dRwD4szKBus67Sm1nEZXung9KlSzNhwgRDxiKEEMIIHkU/4k7YHWwtbKlTqk72L+TnBxs3gqkpjB9vuACLIF1iIjc3bABgd+gQlmXQWTn3lyWEP4QvZ3XKxehEfpTlOWx37959bmHc7du3ywbxQgiRj6QMhzYr2wwzk2z/Px0WLgStFvr1y3h3cpEpPp6exAcH4xNflYQyzUhvJpGi07Jw0xWO3wRKd8z1GEX+kuWEbdq0acyePTvD4z/++CPTp8u2GUIIkV+kzl/LyXBoXBwsWaI+l60Hc+xG8hxwz7AhDBumSXe08/qprdz01/NKCyewrZzLEYr8JssJ25EjR+jWrVuGx7t27crhw4dzFJQQQgjDSZm/1sotByWX1q2DkBBo3Jh0u4NEpoVevUrwhQvE6u04Et6LN95Iv93WP1cA0PvlrrkYncivspywBQYGUrp06QyPlyxZkoCAgBwFJYQQwjDitfGcfXgWDRqal0unKmtmKMrjxQYffCCT33MotZRHWD+atLKmcgadZ9t2HadxZShXp1cuRifyqyxPZnB0dOT27dsZHr916xZ2dnY5CkoIIYRhnHl4hiR9EnVK1sHBMp1dxTNj/364cgXKlIFXXzVsgEVMfGgo9//9FwUNu0Nf4+sMFhtoE2Kp4BBN86pAKZm/JrLRw9amTRuWLVvGo0ePnjn26NEjli9fTuvWrQ0SnBBCiJzZ7LUZUAvmZtuCBerXsWPBQjYez4mbf/yBPjGR81FtiTZzyzD/NQvaw4YJOia+VhesMh7VEkVHlnvYpk2bxvbt22nQoAEfffQR9evXR6PRcP78eX788Ueio6P59NNPjRGrEEKILNDqtay7rA6/vVE3g4lSL+LnB9u2gZmZWntNZJs2Li51scGOkBG8+aZafzg9iTdXYwFQcUhuhSfyuSwnbPXr1+evv/5ixIgRTJ48GU3yXAZFUXB2dmbjxo00biyb0wohRF7bdWsXgTGB1HSuSRPXbO75uXw56HQwYIA6JCqy7e7ff5MQFsbd+Fp4xzVm2wcZNEyKpPfELaDAf8dey80QRT6WrYI8L7/8Mg8ePGDXrl3cvHkTRVGoUaMGXbt2xcrKytAxCiGEyIbVl1YDMKzusNT/XGdJUhIsXao+f+cdA0ZW9Oh1Oq6vWgXAP8Ej6NVLQ/XqGTT22cyFe3q6Ny0FNlLvTqiylLBFR0fTu3dvXn/9dd566y369OljpLCEEELkRFhcGNuub0ODJvvDodu3w8OHUKMGdOhg2ACLGL8DB4i6f58QXVlORXZh34cZt/U/+ysBEdCgaftci0/kf1ladGBra8vp06eNFYsQQggD2XhtIwm6BDpW6oibg1v2LrJokfp17Fgp5ZFD11euBGBH0DAaNDKjbdsMGsY+5PzpIwA06JDBElJRJGVrDpuXl5cxYhFCCGEgqy8mD4fWG5a9C9y8CXv2gJUVDB9uwMiKnqALFwg6f544xZ4DYX359Zfn5L/3/+D8PQWAeo2l4oJ4LMtlPWbMmMGyZcvYv3+/MeIRQgiRQ7dCb3HU5yg25jb0c++XvYssXqx+HTw446WMIlNSetd2hwzCpazN80vZ3VtH86rw2YT+ODhks26eKJSy3MO2du1aypcvT+fOnalXrx7Vq1fH2to6TRuNRsOKFSsMFqQQQojMW3NxDQD9Pfpja2Gb9QvExUFykiGLDXIm6v59fPbuRYcZu0Nf59Nvwdw8g8YRXhB2jk4NHOjUb22uxinyvywnbL/99lvq8wsXLnDhwoVn2kjCJoQQeUOv6FNXhw6vl82hzD//hLAwaNQImmSzHIgA4Prq1aAoHA7rRVIxl+eXsru3jrhEuKPpRA3FLHtlHEShleUhUb1e/8KHTqczRqxCCCFe4MzDM9wLv0dZu7LZ291Aq4XvvlOfjx9v0NiKmoTwcO5s3QrAv6HDeestcHTMoLGiwIONHLsBtYdsxtPTM7fCFAVElhM2IYQQ+dfW61sB6FuzLyaabPwTv349eHtDpUrwRjbLgQgAbm3ciC4+nssxLfFLrMb77z+nceR1iLrBoVtWmJiY0LJly1yLUxQM2e5xjYmJ4fjx4wQEBNC5c2dKlSplyLiEEEJkQ0rC1qdmn6yfnJQEM2aozz///DmTrcSL6JOSuPH77wDsDB5Gnz5QufJzTvDdAsDhOw7Ur+8qCw7EM7LVw7Zo0SLKli1L165dGTZsGFevXgUgKCgIS0tLlqZUxhZCCJFrvIO98Qr2orhlcdpWyKjQ13OsWgV37kD16tK7lkMPdu0iLiCAR9rKXIppzYfPKZQLgM8WErVw/EoobTMs0iaKsiwnbJs2bWL8+PF06NCB5cuXoyhK6jEXFxe6d+/Otm3bDBqkEEKIF0vpXXu5+suYm2axdywhAb76Sn3+5ZfqZu8iWxRFURcbAP8EDqVxYw2tWj3nhBgfCD3DmXvFiE9IpE2bNrkTqChQspywff/993To0IEtW7bwyiuvPHO8cePGXLlyxSDBCSGEyLyt3luBbA6HLl8OPj5QqxYMGmTQuIqa4PPnCb16lVjFkSMRvfnwwxdsFOGrdnLUbdaNf/75hw6yDZhIR5YTtsuXL9O3b98Mj5cpU4bAwMAcBSWEECJr/KP8OeF7AkszS7pV6Za1k+Pi4Jtv1OdffQUmsh4tJ1J613YHD6RUWUsGDHjBCcnz12yrv0rPnj0pLoWKRTqy/LfS1NQUvV6f4fGHDx9iY2OTo6CEEEJkzTZvtZemS+Uu2Fhk8d/gxYvB3x8aNIDn/IdcvFi0ry++yYVyPcOG8O67L1i7kRACgQfRKabMXH8Tb2/vXItVFCxZTtjq1avHrl270j2m1+vZuHEjTaTQohBC5Kony3lkSXT047prX30lm7znkPe6dSh6PcfCXyLRwoUxY15wgt8/oOi4HNOYTz77ijNnzuRKnKLgyXLC9u677/Lvv/8yffp0QkNDATVR8/b25tVXX+Xq1au89957Bg9UCCFE+iLiI9h3dx8mGhNerv5y1k6ePx+CgqBZM+jZ0zgBFhFJMTHc3rQJgP9ChzJiRCa2YfXdCsChB24AskJUZCjLy4AGDRrE5cuX+eabb/gu+X9l3bt3R1EUFEVhxowZ9OjRw+CBCiGESN+OmztI0ifRtkJbXGxcMn+iTqcmbKDWX5PetRzx3bsXbUwMNxMaci/egxf2XSRGwMN/AQ1HrydQvnx53NzcciNUUQBlKWELCgrizp07jBgxgv79+7N27VquX7+OoihUq1aNoUOH0rhxY2PFKoQQIh1/XfsLgP7u/bN2oqcn+PlBlSrQtasRIitafPfuBeBoaHeqV1fL2T2Xz1+gT4BSHThx+iLNmzc3fpCiwMpUwqbX6xk3blyaumstWrRgy5YtuLhk4X9zQgghDComMYb/bv0HQD/3flk7eeVK9eubb0rvWg5p4+N5ePQoAGejOjJocCZOursWgFDHPgQETKZZs2ZGjFAUdJlK2ObPn8/SpUtxdXWlRYsW3Lx5k2PHjvH222+zefNmY8cohBAiA//e+pc4bRzNyzWnnH25zJ8YGgpbt6qJ2vDhRouvqHh0/Di6uDiCzGoRqi1D584vOCHmAQQeAFNLnOoMJzJyLElJSbkRqiigMpWwrV69Gnd3d06cOIGdnR0Ao0eP5rfffiM8PBxHR0djxiiEECID2R4OXb8eEhPVoVCZN5Vjfvv3A3A4oCMaDbRv/4IT7q1Xv5btDRYOWAAWFhbGDFEUcJlaJert7c2bb76ZmqwBTJgwAZ1Ox40bN4wWnBBCiIzFJcWx4+YOIBsJW8pw6IgRBo6q6NHrdPgmJ2ynwjvSqBE4OT3nBEWBe2vU55WG8t577zFv3jzjByoKtEwlbDExMbi6uqZ5LeX7mJgYw0cF6HQ6GjRowMsvq0vUQ0ND6dKlC9WqVaNLly6EhYUZ5b5CCFFQ7L69m+jEaBqWaUil4pUyf+KlS3DuHDg6Qp8+xgqvyAi+cIGE0FASrN3wSahGp04vOCHsAkRcg2Il0JXszMqVK7l582ZuhCoKsEzXYdM8NSE15fsnN383pHnz5uHu7p76/cyZM+nUqRM3b96kU6dOzJw50yj3FUKIgmKTl1rza4D7i/Y+ekpK79prr4GlpYGjKnpShkMvJ3QENC9O2O4m966VH8y16zeJjo6WFaLihTJd1mPnzp08evQo9fvY2Fg0Gg0bN27kwoULadpqNBomTpyY7aB8fX3ZsWMH06ZNY86cOQBs27aNAwcOADB8+HDat2/PrFmzsn0PIYQoyBJ1ifzt/TcA/T2yMByamAhr1dWJjBxphMiKFkVR8NmzB4B/b3eiWDFo3fo5J+i1cP939XmloZzYcgJAVoiKF8p0wrZ+/XrWr1//zOtLlix55rWcJmwffPABs2fPJioqKvW1gIAAypQpA8gG80IIsffOXiISIqhdsjbVS7yo4NcTduyA4GCoXRsaNTJegEVExO3bRPv4gHVxvGPr074DWFk954RHeyH+EdhVgxJNOXlyGU5OTlStWjXXYhYFU6YStv3J3b254Z9//qFkyZI0atQotUctK5YuXcrSpUsBtdCvEEIURsvPLweyOByalARffqk+HzlSaq8ZwP0d6qIPX6v2KJjSpcsLTvD/V/1aYTBoNDg5OdG1a9dnph0J8bRMJWzt2rUzdhypjh49yt9//83OnTuJj48nMjKSN954g1KlSuHv70+ZMmXw9/enZMmS6Z4/ZswYxiTvtiu7LgghCiPvYG+2eG3BwtSCMY1etLv4E+bOVRccVKoEb79ttPiKCr1Wy52tWwH4/WofAAYOfMFJEdfUr05NAJg9e7bR5oKLwiXLm78b23fffYevry/37t3jjz/+oGPHjqxdu5bevXuzatUqAFatWsUrr7ySx5EKIUTe+OHYDygoDKs7jDJ2ZTJ30r178MUX6vOFC8Ha2mjxFRUPDx0iLjAQxakSF0Ib0bKlusvXc6UkbA7uBAcHA88u6hMiPfkuYcvI1KlT8fT0pFq1anh6ejJ16tS8DkkIIXKdf5Q/qy+tRoOGSa0mZe4kRYHx4yEuDgYPhu7djRtkEXHrL7Vo8Vldf0DD0KEvOCEpEuL8wKQYeqsK1KtXj/deuEO8EKosbf6e29q3b0/75HLRJUqUYG/yxrpCCFFUzTs5j0RdIv3c+2V+scGmTbBzJzg4wE8/GTfAIiL20SP8Dx9GY2bGr2dewdwcXn31BSdFeKlf7Wty5NhxHj58SMuWLY0eqygcCkwPmxBCFHUR8REsOrMIgMktJ2fypAhI6cWZORNKlzZSdEXL7S1bUPR64tw6EZHkRM+eUKLEC05KHQ714I8//sDa2ppevXoZPVZROEjCJoQQBcSSs0uITIikXYV2NCuXybpd06aBvz80bw5jsrBAQWRI0eu5s3kzANt91VW6b7yRiRMj1YRNa1ODjRs30qtXL2xsbIwVpihkJGETQogCIEGbwNwTcwGY0mpK5k46dUpdYGBmBkuXgon8k28I/seOEfPwIRYly7HtYnMcHSF5F8XnSx4S3XdVR3BwMIMHDzZqnKJwyddz2IQQQqjWXlqLf7Q/dUrWoXvVTCwa0GrVHjVFgY8+gjp1jB9kEXE7ebHBA6d+KJgwcCAUK5aJE5OHRJu37cXq1dXoLos/RBZIwiaEEPmcXtHz/bHvAZjcanLmykDMmwcXL0LFivD558YNsAiJCw7Gd/9+NKamrDjTF8jkcKg2BmLugcYM+7L1GTq0iVHjFIWP9I8LIUQ+9+/Nf/EO8aa8Q3kG1Rr04hMUBb5XEzwWLJCaawb0YNcuFK0Wy1ptuPagJBUqQKtWmTgx8Aig8OuJUsz7ZSF6vd7YoYpCRhI2IYTI5xaeWQjA+CbjMTc1f/EJXl4QEABlykCPHkaOrmjxS96q0dukGwB9+2ZyauDNBfiGwAcrgtmxY4cUyxVZJgmbEELkY3fC7vDvzX8pZlqMkQ1GZu6klH2Y27eX/UINKCk6msDTp9GYmLDpQlsAMlWVI/ouiu923l2tQas3YfHixZKwiSyThE0IIfKxJWeWoKAwsNZAnK2dM3dSci8QHToYL7AiyP/oUfRaLbbuDThzzRF7e2jdOhMn3lzM5tOw7YzCjBkzqFy5stFjFYWPJGxCCJFPxWvjWXF+BaAOh2aKoqTtYRMG45v8c31k1x5Qd/iysHjBSdo44r2W8d5qaFCnBhMnTjRqjKLwklWiQgiRT228upGQuBAalmlI07JNM3fStWsQHAyurlC1qnEDLEL0Wi3+hw4BsPtOeyCTtdfu/4ElYayfWgOHrn9gZia/dkX2yJ8cIYTIpxacXgDAuMbjMj/nSeavGUXwxYskhIdjU6482/ZXwsQkE+s5FIWEq/MoBrQb8ClUrp8LkYrCSoZEhRAiHzr78Cwn/U7iUMyB1+q8lvkTZTjUKFJWh8a7tScpSUOLFuD8gimF8b4HaTjuIrP/tYYKA3MhSlGYScImhBD5UMom7yPqj8DaPJN11GT+mtH4HTwIwInQ9kDmVod++/kErvlB/Za9wNTSiNGJokASNiGEyGfC4sJYf3k9AGMbj838iTJ/zSii7t8n8s4dzO3s+P1IQ+DF89eunDnAzDVXeKM1dB02KxeiFIWdJGxCCJHPbLi6gThtHJ0qdaKGc43Mn/jff+pXmb9mUH7Jiw3MqrcmIMicSpXAw+P553w6aRz2VjBnUnewqZALUYrCThI2IYTIZzZ7bQbgjbqZ2aTyCevVXjn69jVwREWb/9GjABwPagPA668/Px+Ojo7G89h1RrYDl4ZZ6CEV4jlklagQQuQjYXFh7L+3HxONCS9Xz0zdiGTXr8O5c2Bvn8l6EyIzdAkJBJ4+DcBvh1qi0cBbbz3/HFudD3d/UsDcHsp0z4UoRVEgPWxCCJGP7Ly5E61eS9sKbTO/swE87l3r3x8sZYK7oQSdO4cuPh5tiRoExbrQpQtUrPiCk+7/TmlHKF1nAJgWy4UoRVEgCZsQQuQjW65vAaBvzSwMayoKrFunPn/9dSNEVXSlDIeeCWkJwOjRz2/v6+NDl2E/cOEeUDEL5ViEeAFJ2IQQIp+IS4rj31v/AtCnZp/Mn3jqFNy5A2XKSDkPA0tJ2Pbdb42LC/Tu/fz2G36bw56Lcdg6OEHJ9sYPUBQZkrAJIUQ+4XnHk9ikWBqWaUh5h/KZPzGld23wYDA1NU5wRVBcUBDhN26gM7HEO7Yhb7754r1D169fT+PKULXZEDCRaeLCcCRhE0KIfGLr9a1AFodDtVrYsEF9LsOhBuV/7BgA16KboFUsGDXq+e29r3tx7nogQ1oCFWQ4VBiWJGxCCJEPaPVa/vb+G8hiwrZ3LwQGQo0a0LChkaIrmlKGQ89HtqJdO6he/fntf18xG40GBnUsC84tciFCUZRIwiaEEPnAkQdHCIkLoZpTNTxcXlCV9Ukpw6FDhkixXANS9HoeJfewXYpu9cLFBgCVbe4zthO4NnhDPgthcDLALoQQ+cDaS2sBdbGBJrO/7B89gk2b1OdDhhgpsqLJ/+hREsLCCE4qTZx1Jfr1e/E5w+p5M6w6UH6Q0eMTRY/0sAkhRB57EPGA1RdXY6IxYVTDF0yUetL06RAbqy5dlL1DDUaflMS52bMB8Ax9jVGjNFhZPf+c/3ZsJSbsIZiYQ/F6uRClKGokYRNCiDw268gskvRJDKo1iOolXjBRKsWlS7BiBZiZwfffGzfAIubmhg1E3rlDQKIbeyKH8d57z29/9+5devbuz3d/A9ZuoJFfrcLw5E+VEELkIb9IP5afXw7AtDbTMneSosBHH6lfx4178Wx4kWnxYWFcWrAAgLWPJjNgkAVubs8/Z+7cuZiYaHinM2CdhXIsQmSBJGxCCJGHvj/2PYm6RPq796dWyVqZO+nff2HPHnB0hM8/N2p8Rc3lX34hKTKSa3EtORfdgY8+en77sLAwVqxYwZBeTSnrBNhUyJU4RdEjCZsQQuSRgOgAlpxdAsBnbT/L3ElJSaRmEdOnQ4kSRoqu6Anz9ubWxo0oGlN+eziFTp001K///HOWLFlCTEwMHw6urb5gIz1swjgkYRNCiDzy4/EfidfG07tGb+qXrp+5k5Ytg+vXoUoVGD/eqPEVJYqicG7mTBS9nqPxg/FLqPrC3jWA06dP06VLF+q5JagvSA+bMBIp6yGEEHkgODaYhacXAjC97fTMnRQRAV98oT6fPRuKFTNSdEWP7549BJw6hWLlyOrr4/DwgO7dX3zepk2biI6OhhO91BdkDpswEulhE0KIPDD3xFxikmLoXrU7jV0bZ+6kb7+F4GBo0wb6ZmE3BPFcuoQEziWvtPVMmECM3pGPP35+7VtFUQgODgbA1tYWYu6rB6SHTRiJJGxCCJHLwuLC+Pnkz0AWetfu3oW5c9XnP/4olfQN6PqqVcT4+WFapjprrgygdOkX1yHevXs3bm5uHD9+HPQ6iPNVD1i/YEmpENkkCZsQQuSyn0/+TFRiFB0rdaSlW8sXn6Ao8P77kJgIb7wBTZoYP8giIvzWLa4sXQrAv9op6DFjwoQXjzbPmzeP4sWL06hRI4h9APoksCwJZi+osCtENknCJoQQuSgiPoK5J+cC8HnbTJbkWLMGtm8He3v47jvjBVfEJMXEcOSDD9DFxeHYpjerDjbH2hrGjn3+eSEhIezevZvhw4djYWEBvlvVAy5tjB6zKLokYRNCiFz0y6lfCI8Pp12FdrSr2O7FJ/j5kVpqf948KFfOuAEWEYqicPLzz4m8exeHatXYGK0mz2+9BU5Ozz9369at6HQ6Xn31VfWF+3+qXyvIHqLCeGSVqBBC5JKohCjmHJ8DwOftMtG7pigwerS6OrRnTxg+3MgRFh3ea9fy4L//MLOxodrUuaxtY4WJCUyc+OJzN27cSOXKlWnQoIG62CDkBJhag+tLxg9cFFmSsAkhRC6Zf2o+YfFhtC7fmg4VO7z4hJUr1V0NHB1h6VJZaGAgQefPc/6HHwBo/vXX/LKpIlotDBoElSq9+Py5c+fi5+eHRqOBBxvVF8v2AjMbI0YtijpJ2IQQIhdEJ0bz4/EfAXXumuZFydeDB4+7e375BVxdjRxh0RAfEsKRDz9E0WqpOXw4Ds26smSAeuzjjzN3jZo1a1KzZk31m/sb1K8VBho+WCGeIHPYhBAiFyw8vZCQuBBalGtB58qdn99YUWDUKIiMhFdegddfz50gCzm9TsfRyZOJCwzEpWFD6k+cyLJlEBUF7dtD40yUw5s1axZ79uxRv4m+A6FnwMwWyvQwauxCSMImhBBGFpMYww/H1CG4z9tlondt2TLw9FT3CV2yRIZCDeTy/PkEnDiBZYkStPrxR7SKeWppu0mTXnx+aGgon332GZ6enuoLqcOhvaWchzA6SdiEEMLIFp9ZTFBsEE3LNqVblW7Pb3zv3uPN3RcsgFKljB5fUeB34ABXly5FY2JCqx9+wLpkSf74Q12EW6sW9MhEB9m2bdvQarVPrA6V4VCReyRhE0III4pNiuX7Y+q2Ry+cu6bXw8iREB0NAwbAQEkEDCHa15djn3wCQN333qNU06YoCiSvO3jhNlQpNm7cSMWKFdViuZE3Iew8mNtDmRck4UIYgCRsQghhRGsuriEgJoBGZRrxUrUXlH1YsQL27wcXF1i4UIZCDUCXmMiRDz8kKTKSsh064PHWW4D64718GcqUgddee/F1QkND2bNnD6+++qqadN//Qz1Q9hUwtTTiOxBCJQmbEEIY0fYb2wEY12Tc83vXwsIguReIn39WkzaRY5d+/pnQq1exKVuWFt9+i8bEhLVrYcIE9fjMmS/ehgrg9u3buLq6MnjwYHXv0Dsr1AMVZUGIyB1S1kMIIYwkQZvA/nv7AV48d+2LLyAkBNq1UwuCiRx7eOQIXitXojE1peXs2VjY27NtG7z5proQd/ZsGDYsc9dq0qQJd+7cUZNu///Ugrk2laBMF6O+ByFSSA+bEEIYyTGfY8QmxVK7ZG3K2pfNuOGVK+oYnYmJ2rsmQ6E5FhcczIlPPwWgzvjxuNSvz5496rRAnQ6mTcvcylCAmJgYkpKSMDExURO2W+pm8VQdDRr5NSpyh/xJE0III9l1exfwgt41RYH331eziLFjoW7dXIqu8FL0ek5Mm0Z8SAglmzTBY9Qojh1TS9olJqrDof/7X+avN2/ePMqWLUtERATEPgS/7aAxg8ojjPcmhHiKJGxCCGEkmUrYNm+GffvUHce/+iqXIivcvNeswf/IESwcHGg5axaXLpvy0ksQG6sOh86dm/lOTEVRWLNmDTVr1sTBwQFurwBFB+VeAavSxnwbQqQhCZsQQhhBQHQAFx5dwMrMijYV2qTfKDYWPvxQff7112qhXJEjodeucWHOHEDdJ/RBWCm6doWICOjfX61JbJKF33znz5/n+vXrvPHGG+pig9vL1QNV3zZC9EJkTBYdCCGEEey+vRuAdhXbYWmWQdmH779X9wytVw/GjMnF6AqnpJgYjk6ahF6rpdprr6Gt3JHOrSEoCLp3h3XrwCyLv/XWrl2LhYWFWizXfxfEPgDbylC6k3HehBAZkB42IYQwgt131ISta+Wu6Te4f1+tKQHqQgNT01yKrPA6+913RN27h0O1apR542M6d1Z3MmjTBjZtylz5jidptVp+//13evbsSfHixeHWEvVA1TGy2EDkOulhE0IIA9Mr+tQetm5VM5i/9vHHEB+vlvBo2zYXoyuc7u3cyZ0tWzAtVozan/1A95ctuX0bGjWC7dvB2jrr1zQxMWHDhg3Y2NhArB88/EddbFDpTYPHL8SLSMImhBAGdsL3BIExgZSzL4e7s/uzDdauhb/+AisrdVhU5EhcUBCnZ8wAoN6kqQx5vypXr4KHB/z3Hzg4ZO+6JiYmtE1Jpi/PAEUP5fuDlezvKnKf9OkKIYQBKYrC1D1TARhSe8izuxv8/jsMH64+/+47cHPL5QgLnwtz5pAUHY1ru3asuvgqx45BuXLg6QnOztm7ZkhICO+++y737t0DbRzcmK8eqDbOYHELkRWSsAkhhAFt9trM4QeHcbZ25pM2n6Q9uGEDvPGGusn7jBlq/TWRI8EXL3L3778xMTcnvs1UZs7SYGIC69eDq2v2r7ty5UoWLFhAVFQU3P0NEoLBqTGUbGew2IXIChkSFUIIA0nQJjDJUy2f/1X7r3C0dHx88M8/4fXX1WTtiy/g88/zJshCRNHrOfPttwC4vfom/SeWB+DLL9WFBtml1+tZvHgxrVu3pk4tD/inr3rAY7LsQiHyjPSwCSGEgcw7OY+74Xep5VKL0Y1GPz7w118wZIi6m8H06WrCJnLszrZthF65glXJknx7YDQBAdC+PSTvSJVtnp6e3L59m3HjxoHvVoi+rZbyKNfPEGELkS2SsAkhhAEERAfw9aGvAZjTbQ5mJskDGJs3w+DBjzewnDFDemkMICk6mos//QSAX42P2LnHhhIl1PUcOa2QsmjRIkqWLEm/vn3Ba7b6Ys2PwERKr4i8IwmbEEIYwOf7PycqMYqXqr1E1yrJtde2blXLduh08Mkn6gaWkqwZxJXFi4kPCcGyan0mrugJwKpVULZszq6rKAqurq689957FIs8BSGnoJgzVH4z50ELkQMyh00IIXLoUsAllp9fjqnGlB+6/KC+uG0bvPoqaLUwZQp8840kawYSee8e3mvWgEbDnCufotVqmDgRevbM+bU1Gg0LFy5Uvznwsvq12ngwy0YhNyEMSHrYhBAiBxRF4cNdH6JX9IxrMg53F3e1UmtKsjZpklq+Q5I1gzk3axZ6rZZ7jv04eqcWDRuqP+KcSkxM5MSJEyiKAuFX4eEOMLWC6uNzfnEhckgSNiGEyIF/bvzD3rt7KW5ZnC/afQH//KPuMp6UBB99BLNmSbJmQH4HD/Lw0CGUYnbMOvE+trbwxx9Z33YqPVu3bqVFixYcPHgQrif3lFYeAZYuOb+4EDkkCZsQQmSToihM3jMZgM/bfU6J3YcfJ2sTJ6q7GEiyZjB6rZZzs9VFAH8+fIdIXQkWL4Zq1XJ+ba1Wy08//USlSpVo26QG3Fun7hda88OcX1wIA5A5bEIIkU1nHp7hevB1XO1cGXfNBkb3V+usffAB/PijJGsG9mDXLqLu3SPeyo0dAUMYOFAtbZdTiqIwYcIETpw4wcqVKzEJ2AX6JHB9Ceyq5PwGQhiA9LAJIUQ2bfbaDEC/+EpYvDXmcVHcOXMkWTMwRVG4tnw5AH/6vIUOcz77zDDXnj17NosXL2bKlCm8+eab4L9bPVCmh2FuIIQBSA+bEEJkg6IobPLaBEC/ZUfVF+fNg/fey8OoCq+Hhw8TfuMGWisX9ga9QvfuUKeOYa5dt25dRo8ezbfffqtu8P7IUz1QpothbiCEAUjCJoQQ2XD10SVuht6kRCy08TWBtasNMz4n0nVt2TIAdgQPR6tYMGlSzq8ZHh6Oo6MjPXr0oEeP5N600HPqvqHW5cGues5vIoSByJCoEEJkVUICm79Wk7M+N00x27JNkjUjCjx7lqBz51CK2fO3z0AaNIAOHXJ2zatXr1KlShXWrVuX9kDqcGhXGdYW+YokbEIIkRXR0dCrF5v1VwHoN/RbePnlPA6qcLu2YgUAh+NeI15vw6RJOcul/P39eemll7CwsKB169ZpD6YOh3bN/g2EMAIZEhVCiMwKCYGePbl94yQXW4GdmQ2der2f11EVamHe3jw8eBDMLVl/5Q3Kl1drEmdXdHQ0PXv2JCQkhEOHDlGhQoXHB7UxEHQE0ECpTjmOXQhDkh42IYTIDD8/aNsWTp5kS0tHAF6u2ZtiZgao2CoylNK7dslkAFE6JyZOBLNsdjXodDoGDhzIpUuX2LhxIw0bNkzbIPAQ6BOhRBMo5pTDyIUwLOlhE0KIF7l5E7p0gfv3wcODzT2tIPAs/d3753VkhVq0jw8P/v0XTM1Yfnk4jo4walT2r2diYkLHjh3p06fP40UGT0qZv1ZahkNF/iMJmxBCPI+fH7RpAwEB0KwZt9b9wvG1TbE0s6R71e55HV2hdnXZMhS9nrt2vQjRuvLJO2Brm71rJSYmYmFhwccff5x+A0UB/3/V5zJ/TeRDMiQqhBDP8/HHarLWrh0J/+1g8N53ABjgMQAbC5s8Dq7wCjx7ltubNoGJGQvPv4WNDUyYkL1reXt7U7FiRfbv359xI9+tEOkNlqXAuXn2biSEEUkPmxBCZOTgQXVncUtL+O03Jh6dzln/s1R0rMjP3X/O6+gKLW18PCenTwdgZ/hbPEyswvKFUKZM1q+lKArjx48nNjYWd3f39BvptXDxE/V57elgYp7NyIUwHknYhBAiPVrt4y6dqVNZF3mURWcWYWFqwV+v/kVxq+J5G18hdnnBAqLu3yfMrAob/MbSuzeMHJm9a23YsIG9e/cyf/58SpcunX6jO7+pvWu2VaDK6GzHLYQxScImhBDpWbwYLl+GihW5OvJlxqxpC8DP3X+mkWujPA6u8Aq5fJnrv/2GojHhp5tf4+RiwbJl2au7FhERwcSJE2nUqBFjx45Nv5E2Fi5/oT6v+zWYWmQ/eCGMSBI2IYR4WlAQJA/JRX3/Df23vU5sUixD6w5lTKMxeRxc4aVLTOTE9Okoej27wt/kdlxd/t4AJUtm73qbNm0iICCA7du3Y2pqmn6jG79A3EMo3hAqDMx+8EIYmSw6EEKIp02bBuHhKF06M0rZhneIN7VcarGo5yI0sl2R0VxdtoyImzcJozwb/N9l9Gjo1Sv71xs5ciRXr16lcePG6TdICIWr36nP688EjfxKFPmX/OkUQognnTkDy5eDmRnz323Gn9f+xNbClk0DN8mqUCMK8/bm6tKlAMy/9xXlKlkxZ072rqXX67l16xZAxgsNAK7NhKQIKN0ZynTJ3s2EyCWSsAkhRAq9Xl1ooCicmPgqH12cDcCvvX+lhnONPA6u8NJrtZycPh1Fq2VP2CBuxDdhzZrs11xbvnw5Hh4eXLx4MeNGMT7gnbzSt/7M7N1IiFwkCZsQQqRYswZOnCC4YkkGlj5Mkj6J95u9z6u1crB5pXih66tWEXr1KmH6Mvwe8BGffAItW2bvWoGBgUydOpVWrVpRt27djBte/hL0CVB+EDjJIhKR/0nCJoQQABERMGUKSSYwZKwLPlG+tCjXgtldZud1ZIVamLc3l+bPB2CJz5d41Lfh88+zf73JkycTFRXFwoULM55vGH4F7v4GGjOo93X2byZELpJVokIIATB1KvrAAEaOdsYz/irO1s5sGLABCynzYDSR9++zf/Ro9ImJHAjryy19a86sAYts/sg9PT1ZtWoVU6dOfX6R3BMjQNFD9XFgVzX7b0CIXCQJmxBCHD6MsngxH/cwYa1rMDbmNuwcshM3B7e8jqzQivH3Z99bbxEfEsKVmBasfPQ5a38HD4/sX/PKlSvUqlWLL774IuNG12ZC6BmwLg/1vsn+zYTIZTIkKoQo2uLjYcwYZreCn5rpMTcxZ8ugLTQp2ySvIyu04kNC2DdqFLH+/txJrM+cBz8zaaoFgwbl7LoTJ07k7NmzWFpapt8g7AJcnqE+b74SzO1zdkMhcpEkbEKIou3bb/nV6jpTu4AGDWv6rqFLFSnxYCyJkZHsHzOGqHv3CKAG391ZRJee1nydg6lkx44dY8+ePQAUK1Ys/Ua6RDg+HBQtVBsPpTtm/4ZC5AEZEhVCFF1XrvD3pm8ZPUD99ucePzOodg67eUSGtLGxHHjnHcKuXyfSvCJfXllGhRr2rFsHJtnsPoiNjWX48OEoioKXlxfm5hls3H7lKwi/pO4X2mBW9t+EEHlEEjYhRNGk03F48iAG9dWhN4HpbafzbtN38zqqQkuXmMih998n+MIFkqxK89nFZZjYlWDbNrDPwcjk559/zq1bt9i3b1/GyVrwKXXuGhpo/huYSQFkUfDkuyFRHx8fOnTogLu7O7Vq1WLevHkAhIaG0qVLF6pVq0aXLl0ICwvL40iFEAXZpV8+o1e9a8Sbw5jabzKj/Yy8DqnQ0mu1HJ00iUfHjoFNCaZeXkGY3pUNG6Batexf98SJE/z000+MHTuWDh06pN9IGwcnhoOig5oToWTr7N9QiDyU7xI2MzMzfvzxR7y8vDhx4gQLFizg2rVrzJw5k06dOnHz5k06derEzJlSmVoIkT13rx2lm98sIiyhn31zFvZdLnuEGomi13Ny+nR89+zBxNqeL28s41FiRb7/Hrp2zf51ExISGDlyJGXLlmXWrOcMcV6aDpHXwb4m1JWaa6LgyndDomXKlKFMmTIA2NnZ4e7ujp+fH9u2bePAgQMADB8+nPbt2z//L6kQQqQjMDqArmu68shWoX20M+um7cfUxDSvwyqUFEXh7HffcffvvzG1tGJ+yGJuRtRg2DCYODFn1zY3N2fChAlUqVIF+4zGVAOPwPU56qbuzX8DM6uc3VSIPJTvErYn3bt3j/Pnz9OsWTMCAgJSE7kyZcoQGBiYx9EJIQqaeG08PX5uyi3LWBoEmLBt8hEszTIoASFyRFEULsyZw4316zExN2dbsfkcuVuPZs1gyRLIaYemiYkJ77zzTsYN4oPhxJuAAh6fgHOznN1QiDyW74ZEU0RHR9O/f3/mzp2b8f+e0rF06VIaN25M48aNCQoKMmKEQogCRadj4Zc9OZf0gEph8G/d2dhXlA3djUGflMSJadPw+vVXNKamhHecw+/Hm1O6NGzeDBmVScus4cOH8+uvv2bcIOYB7GkN0bfBsS7UzsFeV0LkE/kyYUtKSqJ///68/vrr9OvXD4BSpUrh7+8PgL+/PyVLlkz33DFjxnDmzBnOnDmDi4tLrsUshMjHIiKI6NuDb5P2AfCL29uUevvDPA6qcEqKjubAuHHc3bYNUysrPL74hcnL1ZpnixeDq2vOrn/y5ElWr16d8X/II7zAsxVEeoNjHejwH5hmUJtNiAIk3yVsiqLw1ltv4e7uzocfPv4HtXfv3qxatQqAVatW8corr+RViEKIguTGDWjWjB+jPQmxhtb2tXnpw0U5H5MTz4gNDGTP8OE8OnaMYk5OdF65kq/WtyMqCvr0AUP8s/3111/j5OTEuHHjnj0YfBI8W0OsL7i0gs4HwapMzm8qRD6Q7xK2o0ePsmbNGvbt20f9+vWpX78+O3fuZOrUqXh6elKtWjU8PT2ZOnVqXocqhMjv/v0XmjYlwNebOa3Uf+5m9l8sK0KNIOL2bXYPGULY9evYVahA13XrOHq3Dps3g60t/PJLzu9x/vx5/vnnHz744APs7OzSHvTfDfs6QWIouL4MHXaDRfGc31SIfCLfLTpo3bo1iqKke2zv3r25HI0QokBSFPj+e5g6FRSFr9+tTIzZHXpV70Wr8q3yOrpCJ/DsWQ69+y6JkZGUqFePdvPnk2TuxPjx6vFvvoFy5XJ+n6+//hoHBwcmTJiQ9sD9DXB8KOiToNIwaLYcTDIooitEAZXvEjYhhMiRuDgYNQrWrwfgzowPWKJZgEav4ZuO3+RxcIXPg127ODZ1KvrERMp17EjL2bMxs7Li0w/B1xcaNSI1ccupt99+m169euHo6Pj4xRsL4cy7gAI1P4IGs9UyHkIUMpKwCSEKDx8f6NsXzp5Vx+HWrOFz/V8kXU5iWL1h1ClVJ68jLFSur17NudmzQVGoNmgQjaZNw8TUlLNnYd48dX/QpUvB1EBl7ro+WWlXUeDyDLiSvENF/ZngPlnmJopCS/4bIoQoHI4ehcaN1WStcmU4fpyLzSux/vJ6LEwtZOspA1L0es7OmsW5WbNAUag/cSKNp0/HxNQUrRbGjAG9Ht5/Hxo2zPn9vL29+fjjjwkJCUkJAM5MUJM1jYk6BOoxRZI1UahJwiaEKPiWLYMOHSAwEDp3htOnoXZtpu2bhoLCO43foaJjxbyOslDQJSRwdNIkvFevxsTMjBYzZ+IxalTqQo4FC+DcOXBzg6++Msw9v/32WxYuXIhOpwNdIhwdAjcXgEkxaP0XVHnLMDcSIh+TIVEhRMGVmAgffqhmCQAffKAuNjAzY9v1bey4uQNbC1s+bfNpnoZZWCRGRXHo3XcJPHMGc1tb2sybR+nmzVOPX70Kn32mPl+wQB2Vzqnbt2+zbt06JkyYQMkSDnCoN/jvAjM7aPc3lGqf85sIUQBID5sQomDy94dOndTMwMICVq6En34iXBvN6L9H02dDHwAmtZxESZv0C22LzEuMiGDfqFEEnjmDValSdF69OjVZUxR1u6kmTSA6Gvr1g169cn7PsLAw+vXrh6WlJZMmTYJzE9VkrZgLdD4gyZooUqSHTQhR8Bw7BgMGqElb2bKwaRM0a8YWry2M3zke/2h/LEwt+KzNZ3zS5pO8jrbASwgPZ9+oUYR5eWFTrhydfv0V27JlAQgJURflbt2qth069HGHZ06NHTuW69ev888//+CasB9uLgITC2i/E5wMMDlOiAJEEjYhRMGhKLBokTr0mZQE7drBhg08slF4988BbPLaBEBLt5Ys67UMDxePvI23EIgPDWXfW28RfuMGtuXL0+nXX7Epo+4esG+fmqA9fAj29upHM2SI4e79ww8/8Oabb9KliSvs6qO+2OhnKNHYcDcRooCQIVEhRMEQFwcjR6pFvZKSYOJElN27+fXhDtwXuLPJaxO2Frb80uMXDo84LMmaAcQFBbH3zTcJv3ED+0qV6LxqFTZlypCYqNYk7txZTdZatoQLFwyTrCUlJbF48WJ0Oh1ubm706NwKDvcDXSxUHApVx+T8JkIUQNLDJoTI/+7fVydGnTsHVlawYgV3ujdjzB8vsfeuugNKj6o9WPzyYso7lM/jYAuH2MBA9o0cSeTduzhUrUrHFSuwcnbm5k01MTtzRq2z9vnn6kIDMwP8NtHr9YwcOZK1a9dSsWJFunfrBifegqgb4FAbmi6W0h2iyJKETQiRv+3ZA4MHq5OlKldGu2kj82L3M31hbeK0cZSwKsG87vMYUmeI7BFqIDH+/uwdOZLoBw9wrF6djitWUKy4EytXwoQJEBMDFSrA2rXQurVh7qkoCu+++y5r167lm2++oXv37nB9Lvj8pa4IbbMJzKwNczMhCiBJ2IQQ+VPKfqCffKJWYe3Rg0vzPuWtA29z5uEZAIbUGcLcbnNxsXHJ42ALj2g/P/aOHEmMry/F3d3puHw5cTgyfDD8+afaZvBgdb7akztE5dSnn37KokWLmDJlCp988gkEHYXzk9SDLX4D++qGu5kQBZAkbEKI/CcqSp2v9tdfAMRP/4Sv28KsPzqg1Wtxs3djUc9F9KzeM48DLVyifXzYM2IEsf7+ONWuTcelSzl5yYE33oAHD9S6agsWqAsNDNmZeefOHebNm8c777zDd999hyYhCI4MBEULNT8Et36Gu5kQBZQkbEKI/OXGDXU/0GvXwN6eI4s+YXTYKq4fvQ7A+Cbj+a7Td9gVs8vjQAuXyPv32TtiBHEBATjXr0/r+Yv5+kc7vvlG7eBs0gTWr4eqVQ1/78qVK3PmzBlq1qyJRtHD0dcg7iG4tFb3CBVCSMImhMhHNm1Se9YiI7nUojILxjVm6U21jlpN55os77WcVuVb5XGQhU/4rVvsHzWKuKAgXBo1ovzHi+jc04bjx9WetE8+gRkzwNzcsPddt24dsbGxjB49Gg8PD3UY/MJUCNgHliWh1QYwMfBNhSigJGETQuS9wECYMAG/f/9kfR1Y09aBy5Z34PYdzEzMmNJqCp+1/QxLM8u8jrTQ8d2/n2NTpqCNiaFkk6ZcrLGA3s2siY9XaxKvXQvt2xv+vlu3bmX48OG0b9+et956CxNFC2fGwe0V6oburf4Aa1fD31iIAkoSNiFE3lEUon7/jc2L3mNN5Wj2fQiKBiACJysnBtUaxPgm46lVslZeR1roKIrCteXLuThvHigKDi16MP3i15xYpSbFQ4fC3Lng5GT4e3t6ejJo0CAaN27M1q1bMUkMhSP9IfAQmFpCi9VQqoPhbyxEASYJmxAi12n1WjxPb2DN+ilstfMjrrP6uoWJBb1q9GJo3aH0qNYDC1OLvA20kNLGxXFy+nTu//svaDQE1Xqf4atGo9VqKFdO3Rf0pZeMc+9jx47Rp08fatSowc6dO7HVPYB9vSD6DliVgbbboEQT49xciAJMEjYhRK5QFIVz/udYc2k1v5/+jUB9JCT33rSxqMbQrh/zaq2BOFo65mmchV3so0cceu89Qq9eRWNpzZ/xs9i6sSMA77wDM2eq20wZy6lTpyhbtiy7d+/GKf4kHB0MSZHg1EhN1qzLGu/mQhRgkrAJIYzqfvh91l1ex9pLa/EK9kp9vUYwDI2rxusf/kZFj5Z5GGHREXThAoffe4/4kBASrNz44up8fOKrUrUqrFgBbdsa7956vR4TExM++OADRo8ahY3vcjj/ESh6KP8qNP9NCuMK8RySsAkhDC48Ppy/rv3FmktrOHT/UOrrLrEaXrukMPSuHY0+nY/G0AW9RIZub9nC6Rkz0CclcUffjFkX5hCrODJpkroC1MrKePe+d+8evXr1YunSpbRo2gibqx/C7WXqwdpfQJ3P1YUGQogMScImhDCIRF0i/936jzWX1rDdezsJugQALE2L0cffgaGegXS5rWDeuw/sXwhlyuRtwEWEXqvl/A8/4L1mDQC7Ql5nXcAkPOqY8+uv0Lixce/v7+9P586dCQkJwcY8CfZ3hcCD6uKC5r9BhUHGDUCIQkISNiFEtimKwkm/k6y5uIYNVzcQEhcCgAYNHSt2ZGhQGfp9sxn78EBwdob182HgQOlVyyWJEREc+fhjHh07hlYx4zf/zzgS8yqfz4CpU8HCyGs6QkND6dq1K48ePWLP1uXUfTgSom/L4gIhskESNiFElt0Ovc3aS2tZe3ktt0Jvpb5eu2RthtYdyhCb5pSbMA2OrFMPDBoEv/wCLrLnZ26JuH2b/e+8S6zfAyK0TszznYtjnUacXwG1cqFKSlRUFN27d+fmzZvsXD2D5mFvq4sLijeEdtvAupzxgxCiEJGETQiRKXpFz5qLa1h6binHfI6lvl7GtgxD6gxhaN2h1HVyRzNvHnzeDeLjoVQpdZfwvn3zMPKix+/QYQ6+/xEkxnAvriYLg3/ho69dee89MDXNnRiKFStG1apVmT6iPh11n6qLC9wGqBu5m9nkThBCFCKSsAkhXuh68HVGbx/NkQdHALAxt6Gfez+G1h1Kx0odMTUxhb17YUJ98EpeCTp8OMyZY5zKqyJDl//aw8UvPsIELScju+JV9RsOelpTpUru3P/ChQvY2NhQrWpV1k9yA6/ZoAC1p0OdL2VxgRDZJAmbECJDibpEZh2ZxdeHvyZRl0hp29J82/FbBtYaiI1Fci/Jgwfw0Ufw11/q91WqqMOfPXrkXeBFkKLAus92oWyZhKlGx+7IEbT87CPmjtHkypRBnU7H999/z+eff06P7t3Z9qkz3FkJGjNo/itUGmr8IIQoxCRhE0Kk67jPcUZvH83VoKsAjG44mlmdZ1HcqrjaID4efvwRvvkG4uLA2hqmTYMPPwRL2fMzNz18CN+8vpPmAVMx1ei4bDeaGTvfp0KF3FnccefOHYYNG8bRo0cZ0L8vi4fFqcmaqTW0+QtcJXkXIqckYRNCpBGVEMW0fdOYf2o+CgrVnKqxtNdS2lds/7jRjh3w/vtw+7b6/cCB8MMP4OaWJzEXVYoCq1bB+k//YZjjJ5ho9Gibj+WbZe9iYpI7ydrx48fp2rUrJiYmrFm5mNdd16AJPgoWTtB+Bzg3z5U4hCjsJGETQqTacWMH7+x4B59IH8xMzJjUchLT207Hyjy5quqtW/DBB2rCBuDhoQ5/duyYZzEXVT4+8PbbEH1sK2NcP8NEo1Bh6HhaTR2XK/dXFAWNRkP9+vUZNGgQn096m/K334Lgy+oK0A67wMEjV2IRoiiQ2Z9CCAKiA3ht02u8/PvL+ET60Ni1MWdGn+HbTt+qyVpMDHz2mVoPYscOdbPJn36CCxckWctligLLl0Pt2hB7fHNqslZnwoRcS9b+/vtvWrduTXR0NFZWViyfM5ny1wdC+GWwrwldjkmyJoSBScImRBGmKAq/XfgN9wXu/HHlD6zNrZnTdQ4n3jpBvdL11Oxg40Zwd1fnqiUmqqs/vb3VnjZz87x+C0XK/fvQrRuMHg2NTP5kjOt0TDQK9T74gDpjxxr9/lFRUYwaNYpXXnmF2NhYQkJCIPQseLaGmHtQohl0OQI2MjQuhKHJkKgQRdTt0Nu8/c/b7L27F4BuVbqxqOciKhWvpDa4dg0mTIB9+9TvGzaE+fOhRYs8irjo0uth6VKYNAmio+GVcn8w0P5/ADT4+GPcR4wwegxHjx5l6NCh3L9/n08++YQvv/wSi9DDsKcPaKOhTDdos0lqrAlhJJKwCVHEaPVafjr+E18c+II4bRwlrEowt/tcXq/zOhqNBiIj1d3Af/4ZtFq1jtq338KoUblXdVWkunsX3noL9u9Xv5/UYR31A74FoOGUKdQcNszoMSiKwqeffopGo+HQoUO0atUKHmyEY2+APhEqDIHmK8HUyHtdCVGEScImRBERFhfGkQdH+OLAF5x/dB6AN+q+wZyuc3CxcVGHP9esUbtxAgLU/T7feQf+9z8oUSKPoy9aFAXu3IEtW+DLL9UphC7OOn4e8hv6PXMAaPTpp9R4/XWjxnH16lWcnZ0pVaoUv//+O3Z2dtjZ2cHNRXB6PKBAjfeh4RwpiCuEkUnCJkQhpCgK98LvceTBEY76HOXIgyOp9dQAKjhUYPHLi+letbv6wuXLMG4cHFF3MqBlS3X1Z8OGeRB90aPXqxtEHDr0+PHwoXpMg54PunvS3mQBMXvUMipNpk+n2uDBRo3pv//+o3///gwYMIBVq1bh6uqqHrg6Ey5+oj6v9y14TCVXKvMKUcRJwiZEIaDVa7n46GKaBM0/2j9NGwtTC5qWbUrXyl2Z2GIitha2EBWlduHMmwc6HZQsCd9/D0OHyi9hI9Jq4eLFx8nZ4cMQEpK2TYkSCoMb7KW1fgH6BzeIAazLlKHBxx9ToXt3o8b3+++/M2zYMGrXrs2sWbPUFxUFLk2Hq98AGmi6GKqOMWocQojHJGETogCKSojihO+J1OTshO8JYpJi0rRxsnKilVsrWpdvTSu3VjRybYSlWfIOBIoCf/yhbin18CGYmMD48fD11+DomPtvqJBLSIDTpx8nZ0ePqrnyk8qWhbZtoW0bhXo2Bwj/ewFhXl7oAevSpak1ZgyV+/bF1MK488Tmz5/Pe++9R9u2bdm2bRsODg7qn5dzE8F7HmhMocVqqDjEqHEIIdKShE2IAsAv0i9N79nFgIvoFX2aNlWdqqZJ0Go418AkvXlF3t5qcrZXXR1K06awaJEMfxpQTAwcP/64B+3kSXUnrydVrQpt2iQnaW2hYkUF/yOHuTx/PnevqsPXVi4ueIweTdUBAzAtVszocUdFRTF79mxeeeUVfv/9dywtLUGvg9Nj4fZyMLGAVhvArY/RYxFCpCUJmxD5jF7RczXwampydtTnKPfC76VpY2ZiRqMyjVKTs1blW1HatvTzLxwTo9ZS++EHSEpSV3/OnKkuQTSRCeM5ERam9pqlJGhnz6rDnk+qXftxctamDaRMCVMUhUfHjuE5bT4hly4BYFmihJqovfoqZrmwL6tOpwPAzs6Oo0ePUqZMGczMzECfBMeHw/3fwdQK2mwB125Gj0cI8SxJ2ITIY3FJcZx+eJojD45w5MERjvseJzw+PE0bOws7Wrq1TE3QmpZtio1FJutdKQps26bu/fnggfraqFHw3Xfg7GzYN1NEBASoQ5spCdqlS+qPOYWJCTRu/DhBa9362YW2iqIQcPIkl+fPJ+i8umq3mJMTHm+9RbVBgzCzssqV95KYmMiwYcNwcnJiwYIFuKXsB6tLgKODwHcbmNlB+3+gZNtciUkI8SxJ2ITIZUExQWl6z84+PEuSPilNGzd7t9TkrHX51tQuWRtTk2zUQLtzRy1+u3On+n2DBrBwITSXDbmz4sGDtCs4vb3THrewUEeWUxK0Fi3U3bsyEnD6NJfnzyfwzBkAijk64j5yJNVfew0za2sjvpO0oqOj6devH56ensyePVutwwegjYVDfeHRbrAoru4LWqJJrsUlhHiWJGxCGJGiKNwMvZnae3bU5yg3Qm6kaaNBQ71S9dIMb5Z3KJ+zG8fHw6xZai9aQoKaPXzzjVpXTYrfPpeiwM2baRO0+/fTtrG2ViufpCRoTZtCZjrEAs+e5fKCBQScPAmAhb097iNGUP311zG3yd0dAoKDg+nZsydnz57l119/ZUTKbglJkXDgZQg6DJYloYMnFK+bq7EJIZ4lCZsQBpSoS+Sc/7nU5Ozog6MExQalaWNlZkXzcs1Te8+al2uOg6WD4YL491+1V+22WrOLN95QS3WUfsEctyJKr1fL0D05xBkQkLaNg0PaBQING2ZtG9WgCxe4vGABj44dA8Dczo6aw4dT4403sLCzM+C7yRy9Xk+3bt24evUqmzdvpnfv3uqBhBDY3x1Cz4B1Oei4B+xr5Hp8QohnScImRA4k6hI5cO8AB+8d5IjPEU75nSJem3Y5YCmbUrQq34rWbq1pXb419UvXx9zUCJum378PH34Imzer33t4qMOf7doZ/l4F3K1b6i4Chw6ptYLDw9MeL1nycXLWtq26YCCrHZOKXk/Q+fNcXbYM/8OHATCzsaHmsGHUHDYMi+eNmRqZiYkJX331Fba2trRL+fMR5w/7ukLEFbCtDB33gm3FPItRCJGWJGxCZFFcUhy7b+9mk9cm/vb+m4iEiDTH3Z3dH5fXKN+KKsWrPJ4bZAzBwepenwsWQGIi2NioxXDffz9r3UBFQECAuk3q0qVqneAU5cureW1KglatWvbqBit6PcGXLvHgv//w8fQk9tEjAMysrKgxdCg1hw+nWB7Wubty5QqXLl1iyJAh9OzZU31Rr1NLdlz8BBLDwN5d7Vmzds2zOIUQz5KETYhMiE6MZufNnWzy2sSOGzvSFKmtXbI2Par2oE35NrR0a0kJ61zadzM6GubMUct0pFRhfe01mD0bypXLnRgKiJgY+PFHdWQ4OlpdxTlkCHTvriZoFSpk/9qpSdquXfjs3p2apIFa8LZir17UHD4cy+LFDfBOss/Ly4tOnTphYWHBK6+8go2NDYSehVPvQOhptVHprtByLVi65GmsQohnScImRAbC48PZ7r2dTV6b2HV7V5qhzkZlGtHfvT/9PfpTvUT13A0sIUHtIvrf/yAoeX5c9+5qL1uDBrkbSz6n1cLKlfD555CSR/XqpZaf8/DI/nVflKS5de1K+W7dcK5bF00+qHF348YNOnbsiEajYc+ePdiYJ8Lpyeom7ihgVRYa/QRuA2RLMiHyKUnYhHhCcGww265vY5PXJvbc2ZOm3EaLci3o796ffu79qFS8Uu4Hp9PB+vVq9nHvnvpa8+bqStD27XM/nnxMUeCff2DKFHVTdYAmTdQetuxO6VMUhZDkJO3Brl35PklLcfv2bTp27IhOp+PA/v3UMD8B2ydBQpC6zVTNiVD7czDP/cUPQojMk4RNFHn+Uf5sub6FTV6bOHjvIDpFndxkojGhfcX29HfvT9+afSlrXzZvAlQU2LEDPv1UXc4IavfQt99C797SI/KUU6dg0iR1QQFApUpqTjtwYNZ/VGmStN27ifX3Tz1mXbo0bl26UL5793yXpD1p+/btxMXFsX/bUjwejoWgI+oBlzbQZCE41s7bAIUQmSIJmyiSHkQ8YNO1TWzy2sQxn2MoqGXqzUzM6Fa5G/3d+/NKzVcoaVMybwM9cgSmTlX3PQJwc4OvvoKhQ6We2lNu34Zp02DDBvV7Jye1M3LsWMjKNpyKohBy+TIP/vvvmSTNqlQpyqf0pNWrl2+TNFDfh0aj4YPxb/FareuU8hkEik6trdbgB6j4hiT7QhQgkrCJIuNW6K3UJO30w9OprxczLUbXKl0Z4DGAXtV7UdwqbyeHA+peR9OmqeN6oO5r9NlnavaRC3tLFiTBwfD112oFk6Qk9cfzwQfqcGhmF2SmJmkpw51PJ2kpPWn5PElL4efnR79+/Vj0eV8aJv5CqbiHoDGBauOh3tdg4ZjXIQohskgSNlFoKYrCtaBrbPJSk7RLAZdSj1mbW/NStZfo796fntV6Ylcsn8zfuXtX7RZat04dCrWxgY8+Uh95WLcrP4qLg3nz1OHOyEi1s+jNN9UOyJTtMJ/nySTNZ/duYh4+TD2WmqR164Zz/foFIklL4e/vT4d2rXjk70vimVNQDSjRVB3+dGqU1+EJIbJJEjZRqCiKwvlH51N70rxDHm/6aF/Mnl7Ve9HfvT/dqnbD2jz39mx8oYAAdeuoxYvVbiJzc3UbqWnT1CquIpVOB2vWwPTp4Ourvta9u7oTV90X7KCUkqT57N7Ng1270iZpJUs+Hu4sYElaigC/u3Rs04SHD0PYNQWa1yoO9WdClVFqD5sQosCShE0UCmcenmHDlQ1s8trE3fC7qa87WTnxSo1X6O/en86VO1PMLAuTmXJDeLhaS23OHLVYmEajzk+bMUOdLS9S6XSwaxd88ok6YgxQv7668rNz5+efG3rtGvd37kw3SXPr2pUKBThJA0BRCLmyls4vv8X9R0n8Oxla9RipJmtSU02IQkESNlGgeQV5MWXPFLbf2J76WimbUvSt2Zf+Hv1pV6GdcbaByglFgZMnYckSdYZ8XJz6eq9eai9bnTp5G18+kpAAe/eq20j9/TcEBqqvly+v/qiGDFGL4GYk/NYtLv70E34HDqS+ZuXiglu3bpTv2hWXBg0KbpIG6kbtd1bDzYXYBHtRszTMfbsS7UavBZeWeR2dEMKAJGETBVJAdABfHviSZeeWoVN02FrYMrL+SAZ4DKClW0tMTfLhCsqICFi7Vi16e+nxfDo6d4YvvoDWrfMutnwkMlLdv37LFti58/EmDgCVK6sjxe+++/y1FzH+/lxesIC727ah6PWYWVlRuW9fynfvXvCTNIDwK3BjAdxbQ1BoDKYm4ORSho2rpkD18WAi/7QLUdjI32pRoMQmxTLn+BxmHZ1FdGI0phpTxjYay5ftv6SUbam8Du9ZigKnT6u9aX/8AbGx6uvOzjBiBIwerW5cWcQFBsK2bbB1K+zZo26JmqJePejbV33UqfP8ShQJ4eFcW74c73Xr0CcmojEzo9rAgdQeOxYrlwI+NKhLBN8tcHMhBKpF5gIjoNNsGxycy3H4xCU0phZ5HKQQwlgkYRMFgk6vY/XF1Xy2/zMeRqlzkHpV78WszrNwd3HP4+jSERmp7kqwZAlcuPD49Q4d4O23oU+frBUHK4Tu3lV70bZsUcvMKWopPDQatbOxb1/1x1S58ouvpY2Lw3vtWq6tWEFScpdc+R49qDdhAnY52Sg0P4j1g1tL1Ud88u4KZrYE2L9Kx/8d4m7AQ/5ZsVCSNSEKOUnYRL63+/ZuJnlOSi3L0ahMI37o+gPtK7bP28DSc+aMmqT9/ru6iADUGmpvvgljxkD1XN53NB9RFHWjhpQk7eLFx8csLKBTJzVJ690bSmWys1Sv1XJn61YuL1hAXPIEt9ItWlB/4kScatUywrvIJYoCgQfUYU/frWrBWwAHD6g2nkdWXenYrTf3H/izc+dO2svWZEIUepKwiXzrUsAlJnlOYvft3QCUdyjPtx2/5bU6r2GSn0oUREWpCdqSJXDu3OPX27VTk7R+/YpssVu9Ho4ff5yk3bnz+JitLfTsqfaivfRS1srMKYqC7969XJw3j8jkixb38KD+xImUaVmAJ9s/sYiAyORNUDVmUP5Vtehtybag0TDq5Zd58OABO3fupF12N0cVQhQokrCJfMcv0o/p+6fz24XfUFBwKObAtDbTmNBsApZm+SjxOXdOXUCwbh1ER6uvOTnB8OFqolazZt7Gl0cSE2HfPjVB27ZNLTGXwsUFXnlF7Unr1Cl7o8KBZ85wfs4cQpK76Gzd3Kj73ntU6N694C4meGIRAdrknlmrMlD1bagyGqxd0zRfuHAhvr6+tCzIyakQIkskYRP5RlRCFLOPzubH4z8Sp43DzMSMcY3HMb3ddJytnfM6PFV0tLp4YMkSdfgzRevW6ty0AQOKZG9adPTjlZ07dqhT+FJUrPh40UDLltnfAjX8xg0uzJ3Lw4MHAbAsUYLaY8dSZcAATC0K4PytdBYRAFCyPVQfB+X6gMnjkjR+fn4sXLiQr776ivLly1O+fPlcD1kIkXckYRN5TqvXsvzccr448AWBMeo8pP7u/fmu03dUK5FPVlBevKgmaWvXPq4z4eio9qaNHg0Feb5UNgUFqbXRtmxRV3YmJDw+VqfO4yStXr2c7TEe8/Ahl+bP5+7ff4OiYGZtjfuIEdQcPhxzG5ucv5HclsEiAioNg2rjwPHZP0s+Pj506NCBwMBAhg0bRo0aNXI5aCFEXpOETeQZRVH458Y/TNkzBa9gdb5Oi3It+KHrD7R0ywdDPbGxj3vTTp16/HrLlmpv2quvgpVV3sWXyxRFXdmZkqQdOaLOUQM1IWvZ8nGSVqVKzu8XFxSE18qV3Fi/Hn1SEiZmZlQdPJjaY8ZgWaJEzm+Qm/Q6CDyo9qals4iASkPBPP39bB88eECHDh0IDg5m9+7dkqwJUURJwiZynU6vY5PXJmYfnc1Z/7MAVClehZmdZ9LfvT+anHTHGMKNG7BoEaxcqRa7BXBwULeMevttqF07b+PLJXo9eHnB4cPq49Chx3t3grrdaZcuaoL2yitQunT27qMoCnEBAYReu0aolxdhyY/YR49S21To2ZN6EyZgm5ld3fMDfRKEnlWTtICDEHxUXVAA6S4iyMju3bsZOXIk0dHReHp60rRp01x6A0KI/EYSNpFr4pLiWHVxFT8c+4HbYbcBcLF2YVqbabzT5B0s8rKOlFYL//wDCxeCp+fj15s1g7FjYeBAsM5Hm8UbQVISnD+vJmaHD6s9aKGhads4OqobM/Ttq67wdHDI2j0UvZ5oHx/Crl9XE7Rr1wi7fp2Ep28EmFlZUapZM+q8+y5O7vmw1t6TdPEQckqdixZ4EIKOgS42bRvbKlDxDag65plFBE+Ki4sjNjaWEiVK4ODgQNmyZVm0aBENGzY08psQQuRnkrAJowuLC2Ph6YX8fOrn1DlqlYtX5uMWH/Nm/TexMs/DYcWAAFi2TB32TOk+srJSN6kcNw4K8S/J2Fg4ceJxD9rx4483YkhRtiy0afP4UavW8/fufJJeqyXy3j3Cnuo5S0pZUfsECwcHiru74+TuTnEPD5w8PLArXz7/rvrUxkDwCTU5CzykPtcnpG1jXxNKtkt+tAXrss+9ZHR0NIsWLeLHH3+kV69eLFu2jGbNmnHixIm873UWQuQ5SdiE0fhE+PDTiZ9YenYpMUlqqYKGZRoypdUU+rv3z7v9PhVF7T5auBA2bVK7lkDdImrcOHUhQfHieRObEYWGqjsKpCRoZ86oHYtPql49bYJWqVLmFgzoEhOJuHVL7TG7do3Q69cJ9/ZGFx//TFsrFxeKe3hQvGZNnJKTM+syZfJ3UpIUCUFHH/eghZwG5akfnmNdNTFLSdAsS2bq0hEREfzyyy/89NNPhIaG0rlzZ4YOHZp6PF//XIQQuUYSNmFwVwOv8v2x71l3eR1avfpLrUvlLkxpNYWOlTrm3S+gqCi1ZtrChWrJfVC7i/r0URO1Tp0y331UAPj6Pk7ODh+GK1fSHjcxgQYN0iZomdlhQBsbS5i3N2FeXmrP2bVrRNy6hf7p7A+wKVs2Ta9Z8Zo1C8aengmhEHTkcQ9a2DlQ9I+Pa0zAqdHj5MylDRRzytatpk+fzi+//ELPnj2ZNm0aLVq0MNCbEEIUJpKwCYM58uAIs47O4p8b/wBgojFhcO3BTGo5iYZl8nBo8do1dRHBqlWPS3KULKmW43j7bSgoE9mfQ1HUtRJPJmh376ZtY2GhTslLSc5atnzx7gKJkZGPEzMvL0KvXSPq3j0UvT5tQ40G+8qVU3vNiru7U7xmTYo5Ohr0fRpNfGBy71lyD1r4ZUB5fFxjBiWaQqnkIU7nlmCRxQl8yQICApgzZw59+vShRYsWTJo0iREjRtCgQQPDvBchRKEkCZvIEb2iZ7v3dmYfm80xn2MAWJpZMrL+SD5q+RGVi2di525jSEpSy+wvWAAHDjx+vXVrtTetf381gymgdDq1NNyTCVryVpqp7OygVavHCVqTJs+v6RsXHJw6zyxlMUC0j88z7TRmZjhWq5aamDm5u+NYo0bBqokW+zC59yy5By1lG6gUJsXAuRm4tFWTNOcWYJaz9+fn58fs2bNZunQpiYmJuLi40KJFC9zc3HArBP9pEEIYlyRsIlsStAmsu7yO7499z/Xg6wAUtyzOu03fZULTCbjY5NGw18OH6iKCpUvV5wA2NmpJjnfegbp18yauHIqPh9OnH6/gPHbscWdhipIlHydnbduqbzW9XQUURSHW3z/tSk0vr9TN059kWqwYjtWrq0Oa7u44eXjgULUqptnZUyovRd973HsWeBCib6c9bmql9pqVbKcmaCWagqnhdqz45JNPmDNnDnq9nqFDhzJ16lSqV69usOsLIQo/SdhElkQmRLL07FJ+OvETD6PUhMjN3o0PW3zIqIajsLWwzf2gFEXtRVu4UK3oqksuSururvamDR2a9foTeSwiQk3KUnrPTp1S9+h8UqVKamKWkqRVq/bsAgFFryfKx0ddCJDcaxZ27RoJ4eHP3NPMxkadb1azZuqcM/tKlTAxK2D/TCgKRN1K24MW+yBtGzM7cGn9eJGAUyMwcFmZ27dvU6lSJUxMTChevDgjR45kypQpVKxY0aD3EUIUDQXsX2KRVx5FP2LeiXksOrOIiAS1mGztkrWZ3HIyg2sPxtzU/AVXMILISFi9Wk3UvJKHtExN1f08x42D9u1ztidSLlEU8PdPm6BdvPh4F4EUdeqkXSBQ9qkqEXqtloi7dx+v1PTyIuz6dbQxMc/cs5ijY2qvWfHkoU07N7f8W0bjefRaiPSGoENqkdqgQxDnn7aNRXF1YUDKIoHi9cHEOP/8Xb16lW+++YYNGzawYcMGBgwYwOTJk41yLyFE0SEJm3ihTdc2MWzrMGKT1CJdbSu0ZXLLybxU7aXcX/EZG6vuLr5hg/o1pWxEmTIwZoy6kODpTCYPKQoEB4OPj7pq09c37fOUx9PVL8zMoGnTx8lZq1bg5JRyTYX44GD8j94g/OZNwr29Cb9xg4jbt9GnlCh5glWpUmlKaBR3d8e6dOmCUS5CFw9xDyHWN+0jzu/x8/hHaVdwAhRzeaLERjtwrK2u7DRWmDod33//PZ6enuzbtw9bW1s+/vhj2rRpY7R7CiGKFknYRIYUReHH4z8y2XMyCgo9q/Xks7af0bxc89wNJD4e/v0X/vwTtm+HJ3uM2reH8ePVvZHMc7eXT69XN0DPKBHz8QE/v7SbomfE0REaNXo8xNmsmbqxgjY+nohbtwg7eJO7yYlZ+M2b6e4MAGBTrpyamLm7qys13d2xcnY27Bs3lKQodSP0uKeSsSdfSwjOxIU0YO2WdojTvqbRelcVRcHLy4u9e/cSFxfH5MmTMTU1ZeXKlVhYWPDFF18wYcIEShS0/U6FEPmaJGwiXVq9lnd3vsuSs0sA+K7Td0xpNSX3emUSEmD3bjVJ27Yt7Qz7Zs1g0CB16NNIq+v0enXV5dOJ2JPP/fyenVeWnuLFoVw5NdRy5Z59Xq4c2FjriXn4kPAbNwi77s257TcJv3GDqPv3ny2hAZjb2eFYvTqO1arhWKNG6vN8sVJTUSAxVE28Yn2fSMie+j5lb83n0ZipOwRYlQXrco8fT35vVQZMjJ+sb9++nT/++IN9+/bxKHmf08aNG6cOd54/fx7rQr59mRAi70jCJp4RmRDJwI0D2XV7F8VMi7G672oG1hpo/BsnJcGePWqStmXL443XQe1+GjhQfeRw0rZOpyZjzxum9PN7vAHC8zg5PT8RK1dOXaT6pMTISHUo88YNwo/e4Fhyr1l6c800pqY4VKmiJmQ1aqgJWvXqebczgKJXa5alGZ58IhlL+V737A4HzzC1BKtyaROxp5Mxy5JGHcrMSFBQEPv27ePgwYPMnTsXCwsLDh48yJ49e+jUqRMdO3akU6dOVKpUKfUcSdaEEMakURRFeXGzgqlx48acOXMmr8MoUHwifOi5vieXAy/jbO3M34P/poWbESuva7Wwf7+apG3enHa38Xr1HidpVatm6nI6HTx6lH4ilvL84cNnt2RKj7NzxomYm5s6Ve55v6P1Wi1R9+4RduOGmpwlP2L9/dNtb+nsrJbQqFEDh2rVKF69OvZVqmCaW/Xi9EnqZP2M5orF+qrzyZ7ekik95vbJydeTyVjZtN9bFM9Xi0K8vb1ZsmQJe/fu5dKlSwDY29tz/PhxPDw8iI2NxcrKqmDM/RNCFEjPy1ukh02kOvvwLL1+74V/tD81StRgx5AdVHGqYvgb6XTqUsgNG9S9PIOCHh/z8FCHOwcOhJo105ym1T5OxjLqHXv48HFVj+dxcXn+MGXZsuoe8JmRsggg/KnELKNFAKbFiuFQtaraa/bEw9Ipe1sbZYo2Lm3y9XQiFusL8QGkqe6fkWLO6Sdjqa+VBXM7470XA4iPj+f48ePs27ePnj170rx5cwIDA1m4cCGtW7fmm2++oVOnTjRq1Aiz5LIm0oMmhMhLkrAJAP72/pvXNr1GbFIs7Su2Z/PAzRS3MuAG6Hq9Wrdiwwb46y8180pRvTr6VwcR0H4Qd61rqcnXTvBZknaY8uHDZ0tdpKdkyecPU5Yt+/yK/8+TsgjgydWZL1oEUPypxMy2fHlM0qtom11JkRlP2k95JKYfXxoaE7As88TQ5BOJWOprrgYtKJub4uLimDt3Lvv27ePIkSPEx8djamqKs7MzzZs3p0WLFoSHh2OZ3T8cQghhRDIkKph3Yh4Td01EQWFYvWEs67UMC0MUEVUUOHkS3e8bUP7ciNkjv9RDwQ6VOVR6EFvMB7IvpB6PAjSZSsZKl352aPLJ566uYIgi/Ir+iUUA3t5E3MzkIoAnHzldBKAokBCS8aT9lNe0US++lon58yfuW5cDy9JGq02W2xRF4dq1a+zbtw8LCwvefvtt9Ho9JUuWxNXVNXUOWrt27bB/0YaqQgiRS2RIVKRLq9cy8b+JzD89H4Cv2n/FZ20/y9IcncREtecrdWjSR0Fz7ixVz26g2YM/KZP4gJS+pHtU4E8G8icDORvRCCIe30ejUUupZZSIlSunJmPGmM6VZhHAEw9tbOwzbTWmpupw5pOrM6tXz3pdM71OHYLMaNJ+SjKmz0RNEFPr50/cty6nDmPmweT93LZp0yY2b96cZiVn586defvttzExMeHu3bvY2eXv4VohhEiPJGxFVHRiNIP/GsyOmzuwMLXg196/8nrd19O0SUhQk7HnraYMCFB7M+pxkUFsYCB/UoU7qdfwpSx/8Sp7nQcRVLkZ5dw0tC4Hg59KylxdjV9GLSeLAFJ6zDK1CECX+LjYa3pzxeL8kifvZ2Kynbnj8xMx63Jg7pCvJu/nlpSVnKdOneKHH35Ao9Hw999/p67kTFnN+eRKTknWhBAFlQyJFhFavZYLjy5w5MERjvoc5dD9QwTGBOJk5cSWgVupoGnDxYukPi5cgDt31FG5xxTK4Ys7XmketUy8cNY/XjgQbVca/5YDiH9lEI4vtaS0q4lRkzFtfDzxwcHEBQcTHxSkfg0OJu7J58HBxIeEoKSzPDR1EcATZTOeWQSg10FCkFpVPy5A/RofAHHJX+Of+JoQkrnALUs+fxWldVkwywd11YxMq9Wi0WgwNTUlJCSEq1evEhERQUREBJGRkURERDBixAhKly7N7t27+eWXX7h//z6XL18G1JWcXl5euLq6EhUVha2trazkFEIUSDIkWgRFJkRywvdEaoJ2wvdE6tZSKZy0HlTZv5VXvqvGk3uBm6KlMnd4xcSLZvZe1DO/RjWdF+Wir2OZFP3szfSoyy7794dBg7Bt04ZqOZxUrygKCeHhxD+VfMUFBaV+Hx8SQlxQEElRmZjDlczWze2JpKwajhWcsXUxxyQx8InE6zJcfzoRCyJTKyhBHXq0cn0i+XpyBeUTxV5NDTDZLg8pikJsbCyRkZHY2tpiZ2dHaGgoe/fuTU20UpKu4cOHU79+fU6ePMmECRNSj0dGRhIbG4unpyedO3dm3759DBz4bM2/tm3bUrp0aWJjY3n48CFly5bltddeo2PHjmlWckoPmhCisJKErZDwifBJTc6OPDjC5cDL6J/aX9EiqhpJt1uhPGgFD1oTE1yeRG7Qnd9pZOVFUzsvquu8cIm4iak2UU3EIp66kbMzuLunfXh4qOOamejV0CUkqInW08nXUz1i8cHB6DNTLA0wMTPD0tkZKxcXLJ1LYOVkj6WDBVb2pljZ6rC0TcTKMgbLYuGY6gIh/gjE/QWhQRCSiWFJADTq/pRWpcGyVPIj+Xnqa8lfizmDiQFXgRqBVqtNkzQ5OTnh5uZGTEwMq1ateibhGjhwIL179+bOnTt06tQp9bguuYbKkiVLGDNmDHfv3n0m4bK1taV169bUr18fS0tLnJycqFixIg4ODjg4OGBvb0/F5GLI7dq1Y+/evdjb26cec3BwSF252adPH/r06ZObPyohhMgXJGErgHR6HVcCr6RJ0HwifZ5qZAb+TeBBK6wf1Mfdpzj1YgLwwIvGNluoybeU1NxDkzLmGZf8SFG+/LOJmbu7mrA9RVEUEiMiUocenxyafDopS4zMxHZEyczt7bFydsbK2RnLEg5YOVhi5WCGpZ2ClU0SltaxWFlGYmEajCY+AOK91d6wJwu7KkBU8iM9xUqkTbYsS4NVOslYMZd8s4IyNjY2NZl6MuFq1KgRAP/73/8IDQ1Nk5D16NGDiRMnkpCQgJOTE7FPLaiYMmUKM2fOJCEhgfHjxwNgbm6emjSlbGKe8jwlkXr6uIeHB5cuXUo9bmdnh+kTva316tXjv//+y/C9lSxZko4dOxr05yWEEIVB/vgNJJ4rJjGGk34nOfrgKAfuHOGk3wlidE8lPvEOFPNpRLkHFaj7wJa2D+NoqL9FLZN1lEia89QFk7+amqo7CHh4pE3KatQAW1t0iYmpw47xwcHE7duXYY9YegVi06MxM0tOwEpgVaI4lsWtsHKwwMoOLG11WFnHYWUVhaVFKKa6IIi/qCZh+qc27dQCkcmPp5k7PtvrZZVOQlbMBQxRviSL/Pz8CA4OTtOLZWdnR+/evQGYMWMGN2/eTHO8adOmLF26FICqVavi/9RCiYEDB7JhwwYA5syZg1arTZNUpfSEWVhYMG7cOOzs7NL0YHl4eADg6OjIo0ePcHBwoFixYs/MBXN2dmb16tUZvjcrKyvq1KljmB+UEEKIVJKw5UP+Uf4cvn+UnVeOcOTBUe7EnkfRpB26sworSUWfsjR+UIyuD2LoFXQPB2XfsxfToZbsr1kT3N1RatYkqVIl4kqWJN7GhriwsMcT869cIf7AgdSELOHJiW0vYG5nl5yIFcequA2WjsWwstNgZadgaR2PlVUMlpbhFDMJQpN4G+KPpb/fZGLy42lmdk8lXRklZCUNXthVURQSEhJSh+Vu3LjB/fv300yK12g0fPDBB4Daw3X48OE0x93c3Dh58iSgJlfHjh1Lc48mTZqkJmyHDh3i3r17qclUhQoVqFChQmrbGTNmoNPp0iRkrq6uqceDg4PT9Go9SaPR8P3332f4Xk1MTChVqlTWf0hCCCGMShK2PKZX9Jx94MXm00fZf/sIXtFHiTS7k6aNRtFQyb84LR6Y0MMngg4PtJSNCgQCH18HiHV2Jq5KFeLKliW+RAnibG2JNzUlLiEhdZgy3ssLXUImanuh1hyzLFECK+cSWBa3xcrREkt7U6zsFKxsk7C0jMHKKhJLixDM9IEQdxl0z9YuAyAh+fEkM5vnzwV7cr6YWfa2BdJqtURFRaUmTaamply9epWLFy+mSagiIyP5+eefMTEx4YcffmDdunVpjpubm6cOI3711VesW7cuzX2cnJxSE7aQkBAiIyNxdHSkQoUKqUlXii+//JKoqKg0PVzFiz/eVWLv3r3PfU+jR49+7vGMkjUhhBAFV4Er6/Hff//x/vvvo9PpGDVqFFOnTs2wbX4r66EocPt+PJtOnGbfzSNcjjhKoMVRdMXC07SzSdDQwleh9X1o4WtCnQAzTBQz4szMiDczI654ceIdHYmzsiJeoyEuIYGE6HRWb2bAzMZGnRvmZIdlcWus7M2wtNNgZavFMnlI0qpYGMVMAtEkBmSukn4KU8vkhCuDuWCWTyRh5rbP+VkpxMfHp5mr5eHhga2tLdeuXWP37t3PlH74+eefKVOmDEuXLmXGjBlEREQQExOTes1Hjx5RqlQpPv/8c/73v/89/nmYmeHg4MC9e/ewtbVl0aJF/Pvvv2kSKgcHByZPnoxGo+HKlSuEhYWlOW5nZ5e6UlEIIYTIjuflLQUqYdPpdFSvXh1PT0/KlStHkyZN+P3331Pn3zwtLxO2xES4dg0Onwtij9dRvIJ3E2xxkIhS3qDRYZ9ghmO8GY5xZlQMN6N2gBmVws1wiTHDQqsmZvHm5ugyWU9KY2JCMScnrEo4YulojZVjMazsNVja6bF6YkjSyiIYM90jSHp6+edzmFg8Zz7YU3PDzOzQK8ozvVcRERHUr1+fsmXLcuvWLVauXPlMwvX999/TpEkTtmzZwsCBA9E+tUr02LFjtGjRgt9++40RI0YAYGNjk5o0bdu2jerVq7Nr1y42btyYZlK8g4MDr776Kra2tjx69IiIiIjU45aWllK3SwghRJ4rNHXYTp06RdWqValcuTIAgwcPZtu2bRkmbLklKCi54OzJSO6c2UFYyL9oza+g2AVgaqqlTLwZNZOTM8fTVbBPMEVD+glClAXwxDx4MysrLJ2LY5U6JGmGlZ0eS5skrKxjsbKMwNIihGKmDzHRXnl+oHogZdqYxuxxT9dTvV9acxduP9ISmWhFRLwZETE6IsOiaNCgAfXr1ycgIIBp06YRGXkjTcI1ffp0XnvtNc6eOUPTpk2fuf26desYMmQIfn5+zJo165mEKiVBq1GjBh9//PEzx2vUqAGQWmLC3t4+3V6tbt260a1btwx/DKVLl6Z06dLP/1kJIYQQ+UiBStj8/Pxwc3NL/b5cuXKpE7nzwpYt8O678J7lK7ha3qIMUObJBo8cMzzX0tYWSxcXrFxd1fphLi7qMKWLC5bOzliG/olV0ErMCeaFBVsV1FWTGtOnaoWlPxyZZFaCAa+PISIipQfsOhERJxk3bhxfffU+4cHB1Kzq8sxt/ve//1G/fn20Wi07d+5Mk0yVK1eOEiVKAFCpUiV++umnZ0o/VKtWDVCLoCYlJWXYq+Xh4cF3332X4du1trbG2jp7c9qEEEKIgqhAJWzpjd4+/Ut/6dKlqeUPgoKCnmlvSA4O6l6b1tVjMAUSTPVEFdOimGmwtbKmdAlX3CrVwqamO5ZVqmDl4oKViwvFihfH5EXznS7vVLvu0KgrHzMzOd+iRKYKtpopCr6+ftja2lKuXLnUhCqljpejoyNr1659JuFycVGTuLJly/Lw4cMMr+/s7Jw6AT89MvwohBBCZE2BStjKlSuHj8/jArG+vr5pyhkAjBkzhjFjxgDqWLAxtWgBN29CKavVJAb7EmpvStWKDQ2TkFQfD1VHJ1fNN+zHpNFoOHv2bIbHzczMeP311zM8LoQQQojcVaAStiZNmnDz5k3u3r1L2bJl+eOPP1i/fn2exWNlpdadBVco60oJQ168mNOL2wghhBCiSChQCZuZmRnz58+nW7du6HQ6Ro4cSa1atfI6LCGEEEIIoypQCRvASy+9xEsvvZTXYQghhBBC5BqTvA5ACCGEEEI8nyRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nEZRFCWvgzAWZ2dnKlasaLDrBQUF4eLiYrDrifxBPtfCST7Xwkk+18JHPtPH7t27R3BwcLrHCnXCZmiNGzfmzJkzeR2GMDD5XAsn+VwLJ/lcCx/5TDNHhkSFEEIIIfI5SdiEEEIIIfI5SdiyYMyYMXkdgjAC+VwLJ/lcCyf5XAsf+UwzR+awCSGEEELkc9LDJoQQQgiRz0nClo7//vuPGjVqULVqVWbOnPnMcUVReO+996hatSp169bl3LlzeRClyIoXfabr1q2jbt261K1bl5YtW3Lx4sU8iFJk1Ys+1xSnT5/G1NSUv/76KxejE9mVmc/1wIED1K9fn1q1atGuXbtcjlBkx4s+14iICHr16kW9evWoVasWK1euzIMo8zFFpKHVapXKlSsrt2/fVhISEpS6desqV69eTdNmx44dSvfu3RW9Xq8cP35cadq0aR5FKzIjM5/p0aNHldDQUEVRFGXnzp3ymRYAmflcU9p16NBB6dGjh7Jx48Y8iFRkRWY+17CwMMXd3V25f/++oiiKEhAQkBehiizIzOf6zTffKJMnT1YURVECAwOV4sWLKwkJCXkRbr4kPWxPOXXqFFWrVqVy5cpYWFgwePBgtm3blqbNtm3bGDZsGBqNhubNmxMeHo6/v38eRSxeJDOfacuWLSlevDgAzZs3x9fXNy9CFVmQmc8V4JdffqF///6ULFkyD6IUWZWZz3X9+vX069eP8uXLA8hnWwBk5nPVaDRERUWhKArR0dE4OTlhZmaWRxHnP5KwPcXPzw83N7fU78uVK4efn1+W24j8I6uf14oVK+jRo0duhCZyILN/V7ds2cLYsWNzOzyRTZn5XG/cuEFYWBjt27enUaNGrF69OrfDFFmUmc/13XffxcvLC1dXV+rUqcO8efMwMZE0JYWkrk9R0lk0q9FostxG5B9Z+bz279/PihUrOHLkiLHDEjmUmc/1gw8+YNasWZiamuZWWCKHMvO5arVazp49y969e4mLi6NFixY0b96c6tWr51aYIosy87nu2rWL+vXrs2/fPm7fvk2XLl1o06YN9vb2uRVmviYJ21PKlSuHj49P6ve+vr64urpmuY3IPzL7eV26dIlRo0bx77//UqJEidwMUWRDZj7XM2fOMHjwYACCg4PZuXMnZmZm9OnTJzdDFVmQ2X+DnZ2dsbGxwcbGhrZt23Lx4kVJ2PKxzHyuK1euZOrUqWg0GqpWrUqlSpW4fv06TZs2ze1w86c8nD+XLyUlJSmVKlVS7ty5kzox8sqVK2na/PPPP2kWHTRp0iSPohWZkZnP9P79+0qVKlWUo0eP5lGUIqsy87k+afjw4bLooADIzOd67do1pWPHjkpSUpISExOj1KpVS7l8+XIeRSwyIzOf69ixY5UvvvhCURRFefTokeLq6qoEBQXlQbT5k/SwPcXMzIz58+fTrVs3dDodI0eOpFatWixevBiAsWPH8tJLL7Fz506qVq2KtbW1LD3O5zLzmX711VeEhIQwbty41HNkM+L8LTOfqyh4MvO5uru70717d+rWrYuJiQmjRo2idu3aeRy5eJ7MfK7Tp0/nzTffpE6dOiiKwqxZs3B2ds7jyPMP2elACCGEECKfk+UXQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQohC6cCBA2g0GjQaDcuXL0+3jUaj4eWXX079vn379mg0GipXrkxiYuIz7b/88ks08Z2VbwAABCRJREFUGs0zNfru3LnDmDFjqFmzJtbW1hQvXhwPDw+GDx/O/v3707StWLFialzpPdauXWuAdy+EKGykcK4QotD74osveP3117GysspU+7t377Jo0SLef//9F7Y9c+YM7dq1w9zcnGHDhlGrVi3i4uK4ceMG27dvx87Ojg4dOqQ5p1y5cnz33XfpXq9Vq1aZilEIUbRIwiaEKNQaN27MmTNnmDt3Lp988skL21tZWVGpUiW+/vprRowY8cKNp2fMmEFsbCznz5+nfv36aY7Nnz+fR48ePXOOg4MDb7zxRpbehxCiaJMhUSFEoTZw4EAaNWrErFmzCAkJeWF7ExMTvvvuO4KDg/n+++9f2P7mzZuUKFHimWQt5VpPb3AthBDZIQmbEKJQ02g0zJo1i4iICL755ptMndO7d29at27NnDlz0u0he1KVKlUICQlh8+bNmY5Jp9MRHByc7kN2CxRCpEcSNiFEodepUye6dOnCwoULuX//fqbOmTVrFrGxsXz55ZfPbffZZ59hbm5O//79qV69OiNHjmTRokV4eXlleM7169dxcXFJ95GZXkAhRNEjCZsQokiYNWsWiYmJTJ8+PVPtW7ZsSZ8+fVixYgXe3t4ZtmvRogVnz55l+PDhREREsHLlSsaNG4eHhwdt2rThzp07z5xTsWJFPD090304ODhk+z0KIQovWXQghCgSGjRowGuvvca6dev4+OOPqVu37gvP+e6779i+fTuffvopmzZtyrBdnTp1+O233wC4f/8+Bw8eZPny5Rw+fJhXXnmFs2fPYmFhkdrexsaGzp075/g9CSGKDulhE0IUGV9//TVmZmZMmTIlU+1r1qzJiBEj2Lx5MydPnszUORUqVGDYsGEcPHiQVq1aceXKFU6dOpWTsIUQQhI2IUTRUalSJd555x3++++/ZwraZmTGjBlYW1szefLkLN1Lo9HQrFkzAPz8/LIcqxBCPEkSNiFEkfLZZ59hb2+f6V42V1dX3n//fQ4dOsTOnTufOe7p6YlWq33m9bi4OHbv3g2Ah4dHzoIWQhR5ModNCFGkODs7M2nSpEwvPgCYMmUKS5cu5fTp088cmzhxIiEhIfTu3Zs6depgbW2Nj48P69ev58aNGwwbNow6deqkOSciIiLDLajq1KlDvXr1svamhBCFniRsQogi58MPP2ThwoX4+/tnqr2DgwPTpk3jww8/fObYnDlz2LZtG0eOHGHTpk2Eh4fj4OBA3bp1mTJlCm+++eYz5/j6+jJ06NB07zVt2jRJ2IQQz9AoUqVRCCGEECJfkzlsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD5nCRsQgghhBD53P8BYuUzmEie17kAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams['figure.figsize'] = (10, 10)\n",
    "\n",
    "ensemblelist = list()\n",
    "lstmlist = list()\n",
    "cfelist = list()\n",
    "nwmlist = list()\n",
    "bpmlist = list()\n",
    "quantiles = np.arange(0, 1, 0.02)\n",
    "for i in quantiles:\n",
    "    ensemblelist.append(perf_df.ensemble.quantile(i))\n",
    "    bpmlist.append(perf_df.bpm.quantile(i))\n",
    "    lstmlist.append(camels_df.LSTM_nnse.quantile(i))\n",
    "    cfelist.append(camels_df.CFE_nnse.quantile(i))\n",
    "    nwmlist.append(camels_df.nwm_nnse.quantile(i))\n",
    "\n",
    "    \n",
    "quantiles = quantiles*100\n",
    "\n",
    "fig, ax = plt.subplots(facecolor=\"white\")\n",
    "\n",
    "ax.plot(lstmlist, quantiles, \"Blue\", linewidth=2, label=\"Long-Short Term Memory\")\n",
    "ax.plot(cfelist, quantiles, \"Red\", linewidth=2, label=\"Conceptual Functional Equivalent\")\n",
    "ax.plot(nwmlist, quantiles, \"Green\", linewidth=2, label=\"National Water Model\")\n",
    "ax.plot(ensemblelist, quantiles, \"Orange\", linewidth=2, label=\"Weighted Ensemble\")\n",
    "ax.plot(bpmlist, quantiles, \"brown\", linewidth=2, label=\"Best Predicted Model\")\n",
    "ax.plot(bamlist, quantiles, \"black\", linestyle=\"--\", label=\"Best Actual Model\")\n",
    "ax.set_xlabel(\"NNSE\", fontsize=18)\n",
    "ax.set_ylabel(\"Percentile\", fontsize=18)\n",
    "ax.legend(fontsize=14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "2eb88e9d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hru_id</th>\n",
       "      <th>ann_P</th>\n",
       "      <th>lon_cen</th>\n",
       "      <th>lat_cen</th>\n",
       "      <th>AREA</th>\n",
       "      <th>elev_mean_x</th>\n",
       "      <th>ave_T</th>\n",
       "      <th>july_T</th>\n",
       "      <th>Perimeter</th>\n",
       "      <th>geometry</th>\n",
       "      <th>gauge_id_x</th>\n",
       "      <th>p_mean</th>\n",
       "      <th>pet_mean</th>\n",
       "      <th>p_seasonality</th>\n",
       "      <th>frac_snow</th>\n",
       "      <th>aridity</th>\n",
       "      <th>high_prec_freq</th>\n",
       "      <th>high_prec_dur</th>\n",
       "      <th>high_prec_timing</th>\n",
       "      <th>low_prec_freq</th>\n",
       "      <th>low_prec_dur</th>\n",
       "      <th>low_prec_timing</th>\n",
       "      <th>gauge_id_y</th>\n",
       "      <th>geol_1st_class</th>\n",
       "      <th>glim_1st_class_frac</th>\n",
       "      <th>geol_2nd_class</th>\n",
       "      <th>glim_2nd_class_frac</th>\n",
       "      <th>carbonate_rocks_frac</th>\n",
       "      <th>geol_porostiy</th>\n",
       "      <th>geol_permeability</th>\n",
       "      <th>gauge_id_x</th>\n",
       "      <th>q_mean</th>\n",
       "      <th>runoff_ratio</th>\n",
       "      <th>slope_fdc</th>\n",
       "      <th>baseflow_index</th>\n",
       "      <th>stream_elas</th>\n",
       "      <th>q5</th>\n",
       "      <th>q95</th>\n",
       "      <th>high_q_freq</th>\n",
       "      <th>high_q_dur</th>\n",
       "      <th>low_q_freq</th>\n",
       "      <th>low_q_dur</th>\n",
       "      <th>zero_q_freq</th>\n",
       "      <th>hfd_mean</th>\n",
       "      <th>gauge_id_y</th>\n",
       "      <th>huc_02</th>\n",
       "      <th>gauge_name</th>\n",
       "      <th>gauge_id_x</th>\n",
       "      <th>soil_depth_pelletier</th>\n",
       "      <th>soil_depth_statsgo</th>\n",
       "      <th>soil_porosity</th>\n",
       "      <th>soil_conductivity</th>\n",
       "      <th>max_water_content</th>\n",
       "      <th>sand_frac</th>\n",
       "      <th>silt_frac</th>\n",
       "      <th>clay_frac</th>\n",
       "      <th>water_frac</th>\n",
       "      <th>organic_frac</th>\n",
       "      <th>other_frac</th>\n",
       "      <th>gauge_id_y</th>\n",
       "      <th>gauge_lat</th>\n",
       "      <th>gauge_lon</th>\n",
       "      <th>elev_mean_y</th>\n",
       "      <th>slope_mean</th>\n",
       "      <th>area_gages2</th>\n",
       "      <th>area_geospa_fabric</th>\n",
       "      <th>gauge_id</th>\n",
       "      <th>frac_forest</th>\n",
       "      <th>lai_max</th>\n",
       "      <th>lai_diff</th>\n",
       "      <th>gvf_max</th>\n",
       "      <th>gvf_diff</th>\n",
       "      <th>dom_land_cover_frac</th>\n",
       "      <th>dom_land_cover</th>\n",
       "      <th>root_depth_50</th>\n",
       "      <th>root_depth_99</th>\n",
       "      <th>LSTM_Unnamed: 0</th>\n",
       "      <th>LSTM_basin</th>\n",
       "      <th>LSTM_NSE</th>\n",
       "      <th>LSTM_fold</th>\n",
       "      <th>LSTM_nnse</th>\n",
       "      <th>CFE_basin_id</th>\n",
       "      <th>CFE_nnse</th>\n",
       "      <th>nwm_Unnamed: 0</th>\n",
       "      <th>nwm_basin_id</th>\n",
       "      <th>nwm_nnse</th>\n",
       "      <th>weight_CFE</th>\n",
       "      <th>weight_LSTM</th>\n",
       "      <th>weight_NWM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1022500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-68.07313</td>\n",
       "      <td>44.79691</td>\n",
       "      <td>6.203873e+08</td>\n",
       "      <td>103.604200</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>312624</td>\n",
       "      <td>POLYGON ((-67.97836 44.61310, -67.97800 44.613...</td>\n",
       "      <td>1022500</td>\n",
       "      <td>3.608126</td>\n",
       "      <td>2.119256</td>\n",
       "      <td>-0.114530</td>\n",
       "      <td>0.245259</td>\n",
       "      <td>0.587356</td>\n",
       "      <td>20.55</td>\n",
       "      <td>1.205279</td>\n",
       "      <td>son</td>\n",
       "      <td>233.65</td>\n",
       "      <td>3.662226</td>\n",
       "      <td>jja</td>\n",
       "      <td>1022500</td>\n",
       "      <td>Acid plutonic rocks</td>\n",
       "      <td>0.590658</td>\n",
       "      <td>Siliciclastic sedimentary rocks</td>\n",
       "      <td>0.164618</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0710</td>\n",
       "      <td>-14.2138</td>\n",
       "      <td>1022500</td>\n",
       "      <td>2.173062</td>\n",
       "      <td>0.602269</td>\n",
       "      <td>1.776280</td>\n",
       "      <td>0.554478</td>\n",
       "      <td>1.702782</td>\n",
       "      <td>0.204734</td>\n",
       "      <td>7.123049</td>\n",
       "      <td>3.90</td>\n",
       "      <td>2.294118</td>\n",
       "      <td>65.15</td>\n",
       "      <td>17.144737</td>\n",
       "      <td>0.0</td>\n",
       "      <td>166.25</td>\n",
       "      <td>1022500</td>\n",
       "      <td>1</td>\n",
       "      <td>Narraguagus River at Cherryfield, Maine</td>\n",
       "      <td>1022500</td>\n",
       "      <td>17.412808</td>\n",
       "      <td>1.491846</td>\n",
       "      <td>0.415905</td>\n",
       "      <td>2.375005</td>\n",
       "      <td>0.626229</td>\n",
       "      <td>59.390156</td>\n",
       "      <td>28.080937</td>\n",
       "      <td>12.037646</td>\n",
       "      <td>1.226913</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.358472</td>\n",
       "      <td>1022500</td>\n",
       "      <td>44.60797</td>\n",
       "      <td>-67.93524</td>\n",
       "      <td>92.68</td>\n",
       "      <td>17.79072</td>\n",
       "      <td>573.60</td>\n",
       "      <td>620.38</td>\n",
       "      <td>1022500</td>\n",
       "      <td>0.9232</td>\n",
       "      <td>4.871392</td>\n",
       "      <td>3.746692</td>\n",
       "      <td>0.863936</td>\n",
       "      <td>0.337712</td>\n",
       "      <td>0.820493</td>\n",
       "      <td>Mixed Forests</td>\n",
       "      <td>0.237435</td>\n",
       "      <td>2.238444</td>\n",
       "      <td>209</td>\n",
       "      <td>1022500</td>\n",
       "      <td>0.743211</td>\n",
       "      <td>fold_4</td>\n",
       "      <td>0.795679</td>\n",
       "      <td>1022500</td>\n",
       "      <td>0.702071</td>\n",
       "      <td>0</td>\n",
       "      <td>1022500</td>\n",
       "      <td>0.778313</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.545250</td>\n",
       "      <td>0.454750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1031500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-69.58119</td>\n",
       "      <td>45.23568</td>\n",
       "      <td>7.665447e+08</td>\n",
       "      <td>304.743490</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>309614</td>\n",
       "      <td>MULTIPOLYGON (((-69.33810 45.12317, -69.33800 ...</td>\n",
       "      <td>1031500</td>\n",
       "      <td>3.522957</td>\n",
       "      <td>2.071324</td>\n",
       "      <td>0.104091</td>\n",
       "      <td>0.291836</td>\n",
       "      <td>0.587950</td>\n",
       "      <td>18.90</td>\n",
       "      <td>1.148936</td>\n",
       "      <td>son</td>\n",
       "      <td>227.35</td>\n",
       "      <td>3.473644</td>\n",
       "      <td>djf</td>\n",
       "      <td>1031500</td>\n",
       "      <td>Siliciclastic sedimentary rocks</td>\n",
       "      <td>0.448928</td>\n",
       "      <td>Metamorphics</td>\n",
       "      <td>0.443863</td>\n",
       "      <td>0.026258</td>\n",
       "      <td>0.0747</td>\n",
       "      <td>-14.8410</td>\n",
       "      <td>1031500</td>\n",
       "      <td>2.030242</td>\n",
       "      <td>0.576289</td>\n",
       "      <td>1.494019</td>\n",
       "      <td>0.445091</td>\n",
       "      <td>1.648693</td>\n",
       "      <td>0.111345</td>\n",
       "      <td>8.010503</td>\n",
       "      <td>18.90</td>\n",
       "      <td>3.286957</td>\n",
       "      <td>94.80</td>\n",
       "      <td>14.697674</td>\n",
       "      <td>0.0</td>\n",
       "      <td>181.00</td>\n",
       "      <td>1031500</td>\n",
       "      <td>1</td>\n",
       "      <td>Piscataquis River near Dover-Foxcroft, Maine</td>\n",
       "      <td>1031500</td>\n",
       "      <td>7.252557</td>\n",
       "      <td>1.279047</td>\n",
       "      <td>0.450236</td>\n",
       "      <td>1.373292</td>\n",
       "      <td>0.559123</td>\n",
       "      <td>35.269030</td>\n",
       "      <td>50.841232</td>\n",
       "      <td>12.654125</td>\n",
       "      <td>0.674594</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1031500</td>\n",
       "      <td>45.17501</td>\n",
       "      <td>-69.31470</td>\n",
       "      <td>247.80</td>\n",
       "      <td>29.56035</td>\n",
       "      <td>769.05</td>\n",
       "      <td>766.53</td>\n",
       "      <td>1031500</td>\n",
       "      <td>0.9548</td>\n",
       "      <td>4.903259</td>\n",
       "      <td>3.990843</td>\n",
       "      <td>0.870668</td>\n",
       "      <td>0.398619</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>Mixed Forests</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>2.400000</td>\n",
       "      <td>466</td>\n",
       "      <td>1031500</td>\n",
       "      <td>0.845839</td>\n",
       "      <td>fold_9</td>\n",
       "      <td>0.866430</td>\n",
       "      <td>1031500</td>\n",
       "      <td>0.657231</td>\n",
       "      <td>1</td>\n",
       "      <td>1031500</td>\n",
       "      <td>0.757158</td>\n",
       "      <td>0.305542</td>\n",
       "      <td>0.376230</td>\n",
       "      <td>0.318228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1047000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-70.16213</td>\n",
       "      <td>44.98744</td>\n",
       "      <td>9.049562e+08</td>\n",
       "      <td>379.779978</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>310157</td>\n",
       "      <td>POLYGON ((-70.10847 45.21669, -70.10858 45.216...</td>\n",
       "      <td>1047000</td>\n",
       "      <td>3.323146</td>\n",
       "      <td>2.090024</td>\n",
       "      <td>0.147776</td>\n",
       "      <td>0.280118</td>\n",
       "      <td>0.628929</td>\n",
       "      <td>20.10</td>\n",
       "      <td>1.165217</td>\n",
       "      <td>son</td>\n",
       "      <td>235.90</td>\n",
       "      <td>3.691706</td>\n",
       "      <td>djf</td>\n",
       "      <td>1047000</td>\n",
       "      <td>Metamorphics</td>\n",
       "      <td>0.308488</td>\n",
       "      <td>Acid plutonic rocks</td>\n",
       "      <td>0.288613</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0522</td>\n",
       "      <td>-14.4819</td>\n",
       "      <td>1047000</td>\n",
       "      <td>2.182870</td>\n",
       "      <td>0.656868</td>\n",
       "      <td>1.415939</td>\n",
       "      <td>0.473465</td>\n",
       "      <td>1.510238</td>\n",
       "      <td>0.196458</td>\n",
       "      <td>8.095148</td>\n",
       "      <td>14.95</td>\n",
       "      <td>2.577586</td>\n",
       "      <td>71.55</td>\n",
       "      <td>12.776786</td>\n",
       "      <td>0.0</td>\n",
       "      <td>184.80</td>\n",
       "      <td>1047000</td>\n",
       "      <td>1</td>\n",
       "      <td>Carrabassett River near North Anson, Maine</td>\n",
       "      <td>1047000</td>\n",
       "      <td>5.359655</td>\n",
       "      <td>1.392779</td>\n",
       "      <td>0.422749</td>\n",
       "      <td>2.615154</td>\n",
       "      <td>0.561181</td>\n",
       "      <td>55.163133</td>\n",
       "      <td>34.185443</td>\n",
       "      <td>10.303622</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.147867</td>\n",
       "      <td>1047000</td>\n",
       "      <td>44.86920</td>\n",
       "      <td>-69.95510</td>\n",
       "      <td>310.38</td>\n",
       "      <td>49.92122</td>\n",
       "      <td>909.10</td>\n",
       "      <td>904.94</td>\n",
       "      <td>1047000</td>\n",
       "      <td>0.9906</td>\n",
       "      <td>5.086811</td>\n",
       "      <td>4.300978</td>\n",
       "      <td>0.891383</td>\n",
       "      <td>0.445473</td>\n",
       "      <td>0.850450</td>\n",
       "      <td>Mixed Forests</td>\n",
       "      <td>0.241027</td>\n",
       "      <td>2.340180</td>\n",
       "      <td>364</td>\n",
       "      <td>1047000</td>\n",
       "      <td>0.800601</td>\n",
       "      <td>fold_7</td>\n",
       "      <td>0.833751</td>\n",
       "      <td>1047000</td>\n",
       "      <td>0.638119</td>\n",
       "      <td>2</td>\n",
       "      <td>1047000</td>\n",
       "      <td>0.525211</td>\n",
       "      <td>0.434267</td>\n",
       "      <td>0.565733</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1052500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-71.17197</td>\n",
       "      <td>44.96168</td>\n",
       "      <td>3.961103e+08</td>\n",
       "      <td>646.073604</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>172588</td>\n",
       "      <td>POLYGON ((-71.10862 45.12730, -71.10825 45.127...</td>\n",
       "      <td>1052500</td>\n",
       "      <td>3.730858</td>\n",
       "      <td>2.096423</td>\n",
       "      <td>0.152097</td>\n",
       "      <td>0.352698</td>\n",
       "      <td>0.561914</td>\n",
       "      <td>13.50</td>\n",
       "      <td>1.129707</td>\n",
       "      <td>jja</td>\n",
       "      <td>193.50</td>\n",
       "      <td>2.896707</td>\n",
       "      <td>mam</td>\n",
       "      <td>1052500</td>\n",
       "      <td>Siliciclastic sedimentary rocks</td>\n",
       "      <td>0.497458</td>\n",
       "      <td>Metamorphics</td>\n",
       "      <td>0.374062</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0711</td>\n",
       "      <td>-15.1658</td>\n",
       "      <td>1052500</td>\n",
       "      <td>2.405105</td>\n",
       "      <td>0.644652</td>\n",
       "      <td>1.301062</td>\n",
       "      <td>0.459700</td>\n",
       "      <td>1.025555</td>\n",
       "      <td>0.305965</td>\n",
       "      <td>8.669019</td>\n",
       "      <td>14.10</td>\n",
       "      <td>2.517857</td>\n",
       "      <td>58.90</td>\n",
       "      <td>7.316770</td>\n",
       "      <td>0.0</td>\n",
       "      <td>197.20</td>\n",
       "      <td>1052500</td>\n",
       "      <td>1</td>\n",
       "      <td>Diamond River near Wentworth Location, NH</td>\n",
       "      <td>1052500</td>\n",
       "      <td>1.301349</td>\n",
       "      <td>1.494807</td>\n",
       "      <td>0.452326</td>\n",
       "      <td>1.262995</td>\n",
       "      <td>0.615538</td>\n",
       "      <td>30.557666</td>\n",
       "      <td>52.614646</td>\n",
       "      <td>11.143326</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.675527</td>\n",
       "      <td>1052500</td>\n",
       "      <td>44.87739</td>\n",
       "      <td>-71.05749</td>\n",
       "      <td>615.70</td>\n",
       "      <td>60.05183</td>\n",
       "      <td>383.82</td>\n",
       "      <td>396.10</td>\n",
       "      <td>1052500</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>4.800830</td>\n",
       "      <td>4.124313</td>\n",
       "      <td>0.880034</td>\n",
       "      <td>0.477328</td>\n",
       "      <td>0.593588</td>\n",
       "      <td>Mixed Forests</td>\n",
       "      <td>0.225615</td>\n",
       "      <td>2.237435</td>\n",
       "      <td>415</td>\n",
       "      <td>1052500</td>\n",
       "      <td>0.696136</td>\n",
       "      <td>fold_8</td>\n",
       "      <td>0.766951</td>\n",
       "      <td>1052500</td>\n",
       "      <td>0.505420</td>\n",
       "      <td>3</td>\n",
       "      <td>1052500</td>\n",
       "      <td>0.696609</td>\n",
       "      <td>0.270973</td>\n",
       "      <td>0.378079</td>\n",
       "      <td>0.350948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1054200</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-71.05872</td>\n",
       "      <td>44.31072</td>\n",
       "      <td>1.813287e+08</td>\n",
       "      <td>629.165746</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>111995</td>\n",
       "      <td>POLYGON ((-70.97999 44.39574, -70.97963 44.395...</td>\n",
       "      <td>1054200</td>\n",
       "      <td>4.067132</td>\n",
       "      <td>2.128355</td>\n",
       "      <td>0.104960</td>\n",
       "      <td>0.299642</td>\n",
       "      <td>0.523306</td>\n",
       "      <td>17.50</td>\n",
       "      <td>1.194539</td>\n",
       "      <td>son</td>\n",
       "      <td>220.30</td>\n",
       "      <td>3.263704</td>\n",
       "      <td>mam</td>\n",
       "      <td>1054200</td>\n",
       "      <td>Metamorphics</td>\n",
       "      <td>0.871443</td>\n",
       "      <td>Siliciclastic sedimentary rocks</td>\n",
       "      <td>0.104295</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0288</td>\n",
       "      <td>-14.2147</td>\n",
       "      <td>1054200</td>\n",
       "      <td>2.731742</td>\n",
       "      <td>0.671663</td>\n",
       "      <td>1.319645</td>\n",
       "      <td>0.437050</td>\n",
       "      <td>1.357808</td>\n",
       "      <td>0.256851</td>\n",
       "      <td>10.095605</td>\n",
       "      <td>16.05</td>\n",
       "      <td>1.888235</td>\n",
       "      <td>82.20</td>\n",
       "      <td>9.963636</td>\n",
       "      <td>0.0</td>\n",
       "      <td>183.45</td>\n",
       "      <td>1054200</td>\n",
       "      <td>1</td>\n",
       "      <td>Wild River at Gilead, Maine</td>\n",
       "      <td>1054200</td>\n",
       "      <td>0.914754</td>\n",
       "      <td>1.477442</td>\n",
       "      <td>0.418461</td>\n",
       "      <td>2.919142</td>\n",
       "      <td>0.594526</td>\n",
       "      <td>58.376169</td>\n",
       "      <td>31.836472</td>\n",
       "      <td>9.851262</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1054200</td>\n",
       "      <td>44.39044</td>\n",
       "      <td>-70.97964</td>\n",
       "      <td>472.31</td>\n",
       "      <td>90.13951</td>\n",
       "      <td>180.98</td>\n",
       "      <td>181.33</td>\n",
       "      <td>1054200</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>5.019827</td>\n",
       "      <td>4.283551</td>\n",
       "      <td>0.886711</td>\n",
       "      <td>0.455084</td>\n",
       "      <td>0.580954</td>\n",
       "      <td>Mixed Forests</td>\n",
       "      <td>0.224857</td>\n",
       "      <td>2.232382</td>\n",
       "      <td>1</td>\n",
       "      <td>1054200</td>\n",
       "      <td>0.609227</td>\n",
       "      <td>fold_0</td>\n",
       "      <td>0.719024</td>\n",
       "      <td>1054200</td>\n",
       "      <td>0.532733</td>\n",
       "      <td>4</td>\n",
       "      <td>1054200</td>\n",
       "      <td>0.547971</td>\n",
       "      <td>0.304790</td>\n",
       "      <td>0.400186</td>\n",
       "      <td>0.295024</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    hru_id  ann_P   lon_cen   lat_cen          AREA  elev_mean_x ave_T july_T  \\\n",
       "0  1022500    0.0 -68.07313  44.79691  6.203873e+08   103.604200  None   None   \n",
       "1  1031500    0.0 -69.58119  45.23568  7.665447e+08   304.743490  None   None   \n",
       "2  1047000    0.0 -70.16213  44.98744  9.049562e+08   379.779978  None   None   \n",
       "3  1052500    0.0 -71.17197  44.96168  3.961103e+08   646.073604  None   None   \n",
       "4  1054200    0.0 -71.05872  44.31072  1.813287e+08   629.165746  None   None   \n",
       "\n",
       "   Perimeter                                           geometry  gauge_id_x  \\\n",
       "0     312624  POLYGON ((-67.97836 44.61310, -67.97800 44.613...     1022500   \n",
       "1     309614  MULTIPOLYGON (((-69.33810 45.12317, -69.33800 ...     1031500   \n",
       "2     310157  POLYGON ((-70.10847 45.21669, -70.10858 45.216...     1047000   \n",
       "3     172588  POLYGON ((-71.10862 45.12730, -71.10825 45.127...     1052500   \n",
       "4     111995  POLYGON ((-70.97999 44.39574, -70.97963 44.395...     1054200   \n",
       "\n",
       "     p_mean  pet_mean  p_seasonality  frac_snow   aridity  high_prec_freq  \\\n",
       "0  3.608126  2.119256      -0.114530   0.245259  0.587356           20.55   \n",
       "1  3.522957  2.071324       0.104091   0.291836  0.587950           18.90   \n",
       "2  3.323146  2.090024       0.147776   0.280118  0.628929           20.10   \n",
       "3  3.730858  2.096423       0.152097   0.352698  0.561914           13.50   \n",
       "4  4.067132  2.128355       0.104960   0.299642  0.523306           17.50   \n",
       "\n",
       "   high_prec_dur high_prec_timing  low_prec_freq  low_prec_dur  \\\n",
       "0       1.205279              son         233.65      3.662226   \n",
       "1       1.148936              son         227.35      3.473644   \n",
       "2       1.165217              son         235.90      3.691706   \n",
       "3       1.129707              jja         193.50      2.896707   \n",
       "4       1.194539              son         220.30      3.263704   \n",
       "\n",
       "  low_prec_timing  gauge_id_y                   geol_1st_class  \\\n",
       "0             jja     1022500              Acid plutonic rocks   \n",
       "1             djf     1031500  Siliciclastic sedimentary rocks   \n",
       "2             djf     1047000                     Metamorphics   \n",
       "3             mam     1052500  Siliciclastic sedimentary rocks   \n",
       "4             mam     1054200                     Metamorphics   \n",
       "\n",
       "   glim_1st_class_frac                   geol_2nd_class  glim_2nd_class_frac  \\\n",
       "0             0.590658  Siliciclastic sedimentary rocks             0.164618   \n",
       "1             0.448928                     Metamorphics             0.443863   \n",
       "2             0.308488              Acid plutonic rocks             0.288613   \n",
       "3             0.497458                     Metamorphics             0.374062   \n",
       "4             0.871443  Siliciclastic sedimentary rocks             0.104295   \n",
       "\n",
       "   carbonate_rocks_frac  geol_porostiy  geol_permeability  gauge_id_x  \\\n",
       "0              0.000000         0.0710           -14.2138     1022500   \n",
       "1              0.026258         0.0747           -14.8410     1031500   \n",
       "2              0.000000         0.0522           -14.4819     1047000   \n",
       "3              0.000000         0.0711           -15.1658     1052500   \n",
       "4              0.000000         0.0288           -14.2147     1054200   \n",
       "\n",
       "     q_mean  runoff_ratio  slope_fdc  baseflow_index  stream_elas        q5  \\\n",
       "0  2.173062      0.602269   1.776280        0.554478     1.702782  0.204734   \n",
       "1  2.030242      0.576289   1.494019        0.445091     1.648693  0.111345   \n",
       "2  2.182870      0.656868   1.415939        0.473465     1.510238  0.196458   \n",
       "3  2.405105      0.644652   1.301062        0.459700     1.025555  0.305965   \n",
       "4  2.731742      0.671663   1.319645        0.437050     1.357808  0.256851   \n",
       "\n",
       "         q95  high_q_freq  high_q_dur  low_q_freq  low_q_dur  zero_q_freq  \\\n",
       "0   7.123049         3.90    2.294118       65.15  17.144737          0.0   \n",
       "1   8.010503        18.90    3.286957       94.80  14.697674          0.0   \n",
       "2   8.095148        14.95    2.577586       71.55  12.776786          0.0   \n",
       "3   8.669019        14.10    2.517857       58.90   7.316770          0.0   \n",
       "4  10.095605        16.05    1.888235       82.20   9.963636          0.0   \n",
       "\n",
       "   hfd_mean  gauge_id_y  huc_02                                    gauge_name  \\\n",
       "0    166.25     1022500       1       Narraguagus River at Cherryfield, Maine   \n",
       "1    181.00     1031500       1  Piscataquis River near Dover-Foxcroft, Maine   \n",
       "2    184.80     1047000       1    Carrabassett River near North Anson, Maine   \n",
       "3    197.20     1052500       1     Diamond River near Wentworth Location, NH   \n",
       "4    183.45     1054200       1                   Wild River at Gilead, Maine   \n",
       "\n",
       "   gauge_id_x  soil_depth_pelletier  soil_depth_statsgo  soil_porosity  \\\n",
       "0     1022500             17.412808            1.491846       0.415905   \n",
       "1     1031500              7.252557            1.279047       0.450236   \n",
       "2     1047000              5.359655            1.392779       0.422749   \n",
       "3     1052500              1.301349            1.494807       0.452326   \n",
       "4     1054200              0.914754            1.477442       0.418461   \n",
       "\n",
       "   soil_conductivity  max_water_content  sand_frac  silt_frac  clay_frac  \\\n",
       "0           2.375005           0.626229  59.390156  28.080937  12.037646   \n",
       "1           1.373292           0.559123  35.269030  50.841232  12.654125   \n",
       "2           2.615154           0.561181  55.163133  34.185443  10.303622   \n",
       "3           1.262995           0.615538  30.557666  52.614646  11.143326   \n",
       "4           2.919142           0.594526  58.376169  31.836472   9.851262   \n",
       "\n",
       "   water_frac  organic_frac  other_frac  gauge_id_y  gauge_lat  gauge_lon  \\\n",
       "0    1.226913           0.0    0.358472     1022500   44.60797  -67.93524   \n",
       "1    0.674594           0.0    0.000000     1031500   45.17501  -69.31470   \n",
       "2    0.000000           0.0    0.147867     1047000   44.86920  -69.95510   \n",
       "3    0.000000           0.0    5.675527     1052500   44.87739  -71.05749   \n",
       "4    0.000000           0.0    0.000000     1054200   44.39044  -70.97964   \n",
       "\n",
       "   elev_mean_y  slope_mean  area_gages2  area_geospa_fabric  gauge_id  \\\n",
       "0        92.68    17.79072       573.60              620.38   1022500   \n",
       "1       247.80    29.56035       769.05              766.53   1031500   \n",
       "2       310.38    49.92122       909.10              904.94   1047000   \n",
       "3       615.70    60.05183       383.82              396.10   1052500   \n",
       "4       472.31    90.13951       180.98              181.33   1054200   \n",
       "\n",
       "   frac_forest   lai_max  lai_diff   gvf_max  gvf_diff  dom_land_cover_frac  \\\n",
       "0       0.9232  4.871392  3.746692  0.863936  0.337712             0.820493   \n",
       "1       0.9548  4.903259  3.990843  0.870668  0.398619             1.000000   \n",
       "2       0.9906  5.086811  4.300978  0.891383  0.445473             0.850450   \n",
       "3       1.0000  4.800830  4.124313  0.880034  0.477328             0.593588   \n",
       "4       1.0000  5.019827  4.283551  0.886711  0.455084             0.580954   \n",
       "\n",
       "      dom_land_cover  root_depth_50  root_depth_99  LSTM_Unnamed: 0  \\\n",
       "0      Mixed Forests       0.237435       2.238444              209   \n",
       "1      Mixed Forests       0.250000       2.400000              466   \n",
       "2      Mixed Forests       0.241027       2.340180              364   \n",
       "3      Mixed Forests       0.225615       2.237435              415   \n",
       "4      Mixed Forests       0.224857       2.232382                1   \n",
       "\n",
       "   LSTM_basin  LSTM_NSE LSTM_fold  LSTM_nnse  CFE_basin_id  CFE_nnse  \\\n",
       "0     1022500  0.743211    fold_4   0.795679       1022500  0.702071   \n",
       "1     1031500  0.845839    fold_9   0.866430       1031500  0.657231   \n",
       "2     1047000  0.800601    fold_7   0.833751       1047000  0.638119   \n",
       "3     1052500  0.696136    fold_8   0.766951       1052500  0.505420   \n",
       "4     1054200  0.609227    fold_0   0.719024       1054200  0.532733   \n",
       "\n",
       "   nwm_Unnamed: 0  nwm_basin_id  nwm_nnse  weight_CFE  weight_LSTM  weight_NWM  \n",
       "0               0       1022500  0.778313    0.000000     0.545250    0.454750  \n",
       "1               1       1031500  0.757158    0.305542     0.376230    0.318228  \n",
       "2               2       1047000  0.525211    0.434267     0.565733    0.000000  \n",
       "3               3       1052500  0.696609    0.270973     0.378079    0.350948  \n",
       "4               4       1054200  0.547971    0.304790     0.400186    0.295024  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "camels_df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "459c52bb",
   "metadata": {},
   "outputs": [],
   "source": [
    "def prettify(ax, title, string):\n",
    "    props = dict(facecolor='white', alpha=0.5)\n",
    "    ax.set_title(title)\n",
    "    ax.set_xlabel(\"Prediction of NNSE\")\n",
    "    ax.set_ylabel(\"NNSE Value\")\n",
    "    lims = [\n",
    "        np.min([ax.get_xlim(), ax.get_ylim()]),  # min of both axes\n",
    "        np.max([ax.get_xlim(), ax.get_ylim()]),  # max of both axes\n",
    "    ]\n",
    "    ax.plot(lims, lims, 'k-', alpha=.5, zorder=0)\n",
    "    ax.text(0.05, 0.95, string, transform=ax.transAxes, fontsize=10, verticalalignment='top', bbox=props, zorder=0)\n",
    "    ax.legend(loc=\"lower right\")\n",
    "    ax.grid()\n",
    "    return ax"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "7dc91d13",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'bin_edges' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Input \u001b[0;32mIn [22]\u001b[0m, in \u001b[0;36m<cell line: 7>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      4\u001b[0m fig, axs \u001b[38;5;241m=\u001b[39m plt\u001b[38;5;241m.\u001b[39msubplots(\u001b[38;5;241m1\u001b[39m, \u001b[38;5;241m3\u001b[39m)\n\u001b[1;32m      6\u001b[0m axs[\u001b[38;5;241m0\u001b[39m]\u001b[38;5;241m.\u001b[39mplot(camels_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124marea_gages2\u001b[39m\u001b[38;5;124m\"\u001b[39m], camels_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mLSTM_nnse\u001b[39m\u001b[38;5;124m\"\u001b[39m], \u001b[38;5;124m'\u001b[39m\u001b[38;5;124m.k\u001b[39m\u001b[38;5;124m'\u001b[39m, markersize\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m2\u001b[39m)\n\u001b[0;32m----> 7\u001b[0m axs[\u001b[38;5;241m0\u001b[39m]\u001b[38;5;241m.\u001b[39mplot(\u001b[43mbin_edges\u001b[49m[\u001b[38;5;241m1\u001b[39m:], bin_means, \u001b[38;5;124m'\u001b[39m\u001b[38;5;124m-r\u001b[39m\u001b[38;5;124m'\u001b[39m)\n\u001b[1;32m      8\u001b[0m axs[\u001b[38;5;241m1\u001b[39m]\u001b[38;5;241m.\u001b[39mplot(camels_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124marea_gages2\u001b[39m\u001b[38;5;124m\"\u001b[39m], camels_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mCFE_nnse\u001b[39m\u001b[38;5;124m\"\u001b[39m], \u001b[38;5;124m'\u001b[39m\u001b[38;5;124m.k\u001b[39m\u001b[38;5;124m'\u001b[39m, markersize\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m2\u001b[39m)\n\u001b[1;32m      9\u001b[0m bin_means, bin_edges, binnumber \u001b[38;5;241m=\u001b[39m stats\u001b[38;5;241m.\u001b[39mbinned_statistic(camels_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124marea_gages2\u001b[39m\u001b[38;5;124m\"\u001b[39m], camels_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mCFE_nnse\u001b[39m\u001b[38;5;124m\"\u001b[39m], statistic\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mmean\u001b[39m\u001b[38;5;124m'\u001b[39m, bins\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m15\u001b[39m)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'bin_edges' is not defined"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from scipy import stats\n",
    "plt.rcParams['figure.figsize'] = (15, 5)\n",
    "\n",
    "fig, axs = plt.subplots(1, 3)\n",
    "\n",
    "axs[0].plot(camels_df[\"area_gages2\"], camels_df[\"LSTM_nnse\"], '.k', markersize=2)\n",
    "axs[0].plot(bin_edges[1:], bin_means, '-r')\n",
    "axs[1].plot(camels_df[\"area_gages2\"], camels_df[\"CFE_nnse\"], '.k', markersize=2)\n",
    "bin_means, bin_edges, binnumber = stats.binned_statistic(camels_df[\"area_gages2\"], camels_df[\"CFE_nnse\"], statistic='mean', bins=15)\n",
    "axs[1].plot(bin_edges[1:], bin_means, '-r')\n",
    "axs[2].plot(camels_df[\"area_gages2\"], camels_df[\"nwm_nnse\"], '.k', markersize=2)\n",
    "bin_means, bin_edges, binnumber = stats.binned_statistic(camels_df[\"area_gages2\"], camels_df[\"nwm_nnse\"], statistic='mean', bins=15)\n",
    "axs[2].plot(bin_edges[1:], bin_means, '-r')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1405e6cd",
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.rcParams['figure.figsize'] = (15, 15)\n",
    "\n",
    "def myplot(ax, name, model):\n",
    "    bin_means, bin_edges, binnumber = stats.binned_statistic(camels_df[name], camels_df[model], statistic='mean', bins=15)\n",
    "    bin_edges_spec = (bin_edges[1:] + bin_edges[:-1]) / 2\n",
    "    ax.plot(camels_df[name], camels_df[model], '.k', markersize=2)\n",
    "    ax.plot(bin_edges_spec, bin_means, '-r')\n",
    "    ax.set_ylabel(\"NNSE\")\n",
    "    ax.set_xlabel(name)\n",
    "    return ax\n",
    "\n",
    "def plotVariables(axs, name):\n",
    "    \n",
    "    axs[0] = myplot(axs[0], name, \"LSTM_nnse\")\n",
    "    axs[1] = myplot(axs[1], name, \"CFE_nnse\")\n",
    "    axs[2] = myplot(axs[2], name, \"nwm_nnse\")\n",
    "    return axs\n",
    "\n",
    "\n",
    "fig, axs = plt.subplots(4,3)\n",
    "axs[0, :] = plotVariables(axs[0, :], \"area_gages2\")\n",
    "axs[1, :] = plotVariables(axs[1, :], \"frac_snow\")\n",
    "axs[2, :] = plotVariables(axs[2, :], \"p_mean\")\n",
    "axs[3, :] = plotVariables(axs[3, :], \"aridity\")\n",
    "\n",
    "cols = [\"LSTM\", \"CFE\", \"NWM\"]\n",
    "rows = [\"Basin Area\", \"Snow Fraction\", \"Precipitation Mean\", \"Aridity\"]\n",
    "\n",
    "for ax, col in zip(axs[0], cols):\n",
    "    ax.set_title(col)\n",
    "\n",
    "for ax, row in zip(axs[:,0], rows):\n",
    "    ax.set_ylabel(row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9e8eed8f",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, axs = plt.subplots(4,3)\n",
    "axs[0, :] = plotVariables(axs[0, :], \"low_prec_dur\")\n",
    "axs[1, :] = plotVariables(axs[1, :], \"elev_mean_x\")\n",
    "axs[2, :] = plotVariables(axs[2, :], \"silt_frac\")\n",
    "axs[3, :] = plotVariables(axs[3, :], \"sand_frac\")\n",
    "\n",
    "cols = [\"LSTM\", \"CFE\", \"NWM\"]\n",
    "rows = [\"Low Precipitation\", \"Elevation Mean\", \"Silt Fraction\", \"Sand Fraction\"]\n",
    "\n",
    "for ax, col in zip(axs[0], cols):\n",
    "    ax.set_title(col)\n",
    "\n",
    "for ax, row in zip(axs[:,0], rows):\n",
    "    ax.set_ylabel(row)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}