dual-input model with wv42 wavelet data!!!!!.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "a5cc3604",
   "metadata": {},
   "outputs": [],
   "source": [
    "################################################################################################################################\n",
    "\n",
    "                              ####### THIS NOTEBOOK IS FOR TESTING PURPOSES ONLY!!!!! #######\n",
    "                        ####### TO TRAIN A WORKING CNN MODEL, REFER TO OTHER NOTEBOOKS!!!!! #######\n",
    "\n",
    "################################################################################################################################"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "cdb670f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "## importing all libraries for our models\n",
    "\n",
    "# pandas\n",
    "import pandas as pd\n",
    "\n",
    "# tensorflow\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "\n",
    "# tensorflow.keras\n",
    "from keras import Input\n",
    "from keras.models import Sequential, Model\n",
    "from keras.layers import Dense, Dropout, Flatten \n",
    "from keras.layers import Conv2D, Conv1D, MaxPooling2D\n",
    "from keras.layers import AveragePooling2D, AveragePooling1D\n",
    "from keras.layers import GlobalMaxPooling1D, GlobalAveragePooling2D\n",
    "from keras.layers import BatchNormalization, LSTM\n",
    "from keras.layers import Concatenate, concatenate \n",
    "# Concatenate returns Concatenate object (layers), concatenate returns Tensor object (arrays)\n",
    "from keras.wrappers.scikit_learn import KerasClassifier\n",
    "from tensorflow.keras.optimizers import RMSprop\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "from keras import backend as K\n",
    "from keras.constraints import max_norm, non_neg, unit_norm, min_max_norm\n",
    "#from tensorflow.keras.utils import plot_model\n",
    "\n",
    "\n",
    "# numpy\n",
    "import numpy as np\n",
    "from numpy import random\n",
    "\n",
    "# sklearn\n",
    "import sklearn\n",
    "\n",
    "# sklearn for pipeline mixins\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.base import BaseEstimator, TransformerMixin\n",
    "\n",
    "# sklearn for data splitting\n",
    "from sklearn.model_selection import train_test_split, cross_val_score, GridSearchCV\n",
    "\n",
    "# sklearn for models\n",
    "from sklearn.svm import SVC  \n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn import neighbors\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.tree import export_graphviz\n",
    "from sklearn.preprocessing import normalize\n",
    "\n",
    "# sklearn for model evaluation\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import recall_score\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import f1_score, roc_curve, auc, RocCurveDisplay\n",
    "import sklearn.metrics.cluster as smc\n",
    "\n",
    "# matplotlib.pyplot for plotting\n",
    "from matplotlib import pyplot\n",
    "import matplotlib.pyplot as plt \n",
    "import os\n",
    "\n",
    "# misc\n",
    "import itertools\n",
    "from IPython.display import display\n",
    "import matplotlib.pyplot as plt\n",
    "import pydot\n",
    "import graphviz\n",
    "from PIL import Image\n",
    "import numpy as np\n",
    "from numpy import random\n",
    "import glob\n",
    "import rasterio as rio\n",
    "import scipy.io as sio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "b1e6c809",
   "metadata": {},
   "outputs": [],
   "source": [
    "def scipyio_read_file(files):\n",
    "    for i, f in enumerate(files):\n",
    "        if i == 0:\n",
    "            im = sio.loadmat(f)\n",
    "            hold_matrix = im[\"imgs\"]\n",
    "        else:\n",
    "            im = sio.loadmat(f)\n",
    "            imarray = im[\"imgs\"]\n",
    "            hold_matrix = np.concatenate((hold_matrix, imarray), axis = 1)\n",
    "    hold_matrix = np.swapaxes(hold_matrix, 1, 3)\n",
    "    hold_matrix = np.swapaxes(hold_matrix, 2, 1)\n",
    "    return hold_matrix\n",
    "\n",
    "def process_lake_images(data):\n",
    "    # key = index of the lakes, value = list of image data sorted by date for that lake\n",
    "    result = {} # reset dict every time function is called\n",
    "    res = data.index.values\n",
    "    row = None\n",
    "    for it in np.unique(res):\n",
    "        row = data.loc[it]\n",
    "        row = row.sort_values(by=['date'])\n",
    "        result[it] = np.asarray(row).reshape(1,row.shape[0], row.shape[1])\n",
    "    for i in sorted(result.keys()):\n",
    "        if i == 0:\n",
    "            full_feat_data = result[i]\n",
    "        else:\n",
    "            full_feat_data = np.append(full_feat_data, result[i], axis = 0)\n",
    "    \n",
    "    return full_feat_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "c39e17ec",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(252, 36, 36, 144) (252,)\n",
      "(251, 36, 36, 144) (251,)\n"
     ]
    }
   ],
   "source": [
    "# process and read image data (scipy.io/.mat)\n",
    "\n",
    "therm_db4_files = glob.glob('/home/fun/simulation/data/thermokarst_lakes_preprocessed/wavelet-decomposed-py/wv42/*')\n",
    "therm_img = scipyio_read_file(therm_db4_files)\n",
    "\n",
    "therm_img_label = np.ones(therm_img.shape[0])\n",
    "print(therm_img.shape, therm_img_label.shape)\n",
    "\n",
    "glac_db4_files = glob.glob('/home/fun/simulation/data/glacial_lakes_preprocessed/wavelet-decomposed-py/wv42/*')\n",
    "glac_img = scipyio_read_file(glac_db4_files)\n",
    "\n",
    "glac_img_label = np.zeros(glac_img.shape[0])\n",
    "print(glac_img.shape, glac_img_label.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "7a0f80bc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(252, 17) (252,)\n",
      "(251, 17) (251,)\n"
     ]
    }
   ],
   "source": [
    "# process and read time-series data (pandas/.pkl)\n",
    "\n",
    "therm_feat_data = pd.read_pickle(\"/home/fun/simulation/data/thermokarst_lakes_preprocessed/non-image data.pkl\")\n",
    "glac_feat_data = pd.read_pickle(\"/home/fun/simulation/data/glacial_lakes_preprocessed/non-image data.pkl\")\n",
    "'''\n",
    "therm_feat_data['date'] = therm_feat_data['date'].astype('int')\n",
    "therm_feat_data['date'] = pd.to_datetime(therm_feat_data['date'], format='%Y%m%d')  \n",
    "\n",
    "therm_feat_sort = therm_feat.sort_index()\n",
    "\n",
    "glac_feat_data['date'] = glac_feat_data['date'].astype('int')\n",
    "glac_feat_data['date'] = pd.to_datetime(glac_feat_data['date'], format='%Y%m%d')  \n",
    "\n",
    "glac_feat_sort = glac_feat.sort_index()\n",
    "\n",
    "therm_feat_data = process_lake_images(therm_feat_sort)\n",
    "glac_feat_data = process_lake_images(glac_feat_sort)\n",
    "\n",
    "therm_feat_data = np.delete(therm_feat_data, [0, 1, 2, 3], axis = 2)\n",
    "glac_feat_data = np.delete(glac_feat_data, [0, 1, 2, 3], axis = 2)\n",
    "'''\n",
    "therm_feat_label = np.ones(therm_feat_data.shape[0])\n",
    "print(therm_feat_data.shape, therm_feat_label.shape)\n",
    "\n",
    "glac_feat_label = np.zeros(glac_feat_data.shape[0])\n",
    "print(glac_feat_data.shape, glac_feat_label.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "3592090a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(503, 36, 36, 144)\n",
      "(503, 17)\n"
     ]
    }
   ],
   "source": [
    "# img\n",
    "non_shuf_spatial_data = np.append(therm_img, glac_img, axis = 0)\n",
    "non_shuf_spatial_data = non_shuf_spatial_data.astype(\"float32\")\n",
    "\n",
    "non_shuf_spatial_label = np.append(therm_img_label, glac_img_label, axis = 0)\n",
    "\n",
    "non_shuf_spatial_label = non_shuf_spatial_label.reshape(non_shuf_spatial_label.shape[0], 1)\n",
    "print(non_shuf_spatial_data.shape)\n",
    "\n",
    "# time\n",
    "non_shuf_feat_data = np.append(therm_feat_data, glac_feat_data, axis = 0)\n",
    "non_shuf_feat_data = non_shuf_feat_data.astype(\"float32\")\n",
    "\n",
    "non_shuf_feat_label = np.append(therm_feat_label, glac_feat_label, axis = 0)\n",
    "\n",
    "non_shuf_feat_label = non_shuf_feat_label.reshape(non_shuf_feat_label.shape[0], 1)\n",
    "print(non_shuf_feat_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "22f9e571",
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(non_shuf_spatial_data.shape[0]):\n",
    "    for j in range(non_shuf_spatial_data.shape[3]):\n",
    "        scaler = StandardScaler()\n",
    "        scaler.fit(non_shuf_spatial_data[i, :, :, j])\n",
    "        non_shuf_spatial_data[i, :, :, j] = scaler.transform(non_shuf_spatial_data[i, :, :, j])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "aba9f6ec",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "numpy.ndarray"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# normalize data\n",
    "\n",
    "scaler = StandardScaler()\n",
    "\n",
    "scaler.fit(non_shuf_feat_data)\n",
    "\n",
    "non_shuf_feat_scaler = scaler.transform(non_shuf_feat_data) # see scikit-learn documentation for details\n",
    "\n",
    "type(non_shuf_feat_scaler)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "61124a78",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(352, 1) (151, 1)\n",
      "177.0 75.0\n",
      "(352, 1) (151, 1)\n",
      "177.0 75.0\n"
     ]
    }
   ],
   "source": [
    "# img\n",
    "X_train_spatial, X_test_spatial, y_train_spatial, y_test_spatial = train_test_split(\n",
    "    non_shuf_spatial_data, non_shuf_spatial_label, train_size = 0.7, random_state = 3928983)\n",
    "\n",
    "train_spatial_ones = np.sum(y_train_spatial)\n",
    "test_spatial_ones = np.sum(y_test_spatial)\n",
    "\n",
    "print(y_train_spatial.shape, y_test_spatial.shape)\n",
    "print(train_spatial_ones, test_spatial_ones)\n",
    "\n",
    "# time\n",
    "X_train_feat, X_test_feat, y_train_feat, y_test_feat = train_test_split(\n",
    "    non_shuf_feat_scaler, non_shuf_feat_label, train_size = 0.7, random_state = 3928983)\n",
    "\n",
    "train_feat_ones = np.sum(y_train_feat)\n",
    "test_feat_ones = np.sum(y_test_feat)\n",
    "\n",
    "print(y_train_feat.shape, y_test_feat.shape)\n",
    "print(train_feat_ones, test_feat_ones)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "3ec82fe2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            [(None, 36, 36, 144) 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d (Conv2D)                 (None, 36, 36, 9)    104985      input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d (AveragePooli (None, 12, 12, 9)    0           conv2d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization (BatchNorma (None, 12, 12, 9)    36          average_pooling2d[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_1 (Conv2D)               (None, 12, 12, 9)    6570        batch_normalization[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_1 (AveragePoo (None, 4, 4, 9)      0           conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNor (None, 4, 4, 9)      36          average_pooling2d_1[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "flatten (Flatten)               (None, 144)          0           batch_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "input_2 (InputLayer)            [(None, 17)]         0                                            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate (Concatenate)       (None, 161)          0           flatten[0][0]                    \n",
      "                                                                 input_2[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense (Dense)                   (None, 128)          20736       concatenate[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "dropout (Dropout)               (None, 128)          0           dense[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "dense_1 (Dense)                 (None, 32)           4128        dropout[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dropout_1 (Dropout)             (None, 32)           0           dense_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_2 (Dense)                 (None, 1)            33          dropout_1[0][0]                  \n",
      "==================================================================================================\n",
      "Total params: 136,524\n",
      "Trainable params: 136,488\n",
      "Non-trainable params: 36\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# feature extraction from spatial data\n",
    "spatial_inputs = Input(shape = (36, 36, 144))\n",
    "\n",
    "spatial_conv1 = Conv2D(filters = 9, \n",
    "               kernel_size = (9, 9), \n",
    "               activation = 'relu', \n",
    "               padding = \"same\")(spatial_inputs)\n",
    "\n",
    "spatial_pool1 = AveragePooling2D(pool_size = (3,3), \n",
    "                         strides = 3)(spatial_conv1)\n",
    "\n",
    "spatial_batch1 = BatchNormalization()(spatial_pool1)\n",
    "\n",
    "spatial_conv2 = Conv2D(filters = 9, \n",
    "               kernel_size = (9, 9), \n",
    "               activation = 'relu', \n",
    "               padding = \"same\")(spatial_batch1)\n",
    "\n",
    "spatial_pool2 = AveragePooling2D(pool_size = (3,3), \n",
    "                         strides = 3)(spatial_conv2)\n",
    "\n",
    "spatial_batch2 = BatchNormalization()(spatial_pool2)\n",
    "\n",
    "spatial_flat = Flatten()(spatial_batch2)\n",
    "\n",
    "# feature extraction from non-img feature data\n",
    "feat_inputs = Input(shape = (17))\n",
    "\n",
    "# concatenate both feature layers and define output layer after some dense layers\n",
    "\n",
    "concat = concatenate([spatial_flat, feat_inputs])\n",
    "\n",
    "dense1 = Dense(128, activation = 'relu')(concat)\n",
    "dense1dropout = Dropout(0.5)(dense1)\n",
    "\n",
    "dense2 = Dense(32, activation = 'relu')(dense1dropout)\n",
    "dense2dropout = Dropout(0.5)(dense2)\n",
    "\n",
    "output = Dense(1, activation = 'sigmoid')(dense2dropout)\n",
    "\n",
    "# create model with two inputs\n",
    "model = Model([spatial_inputs, feat_inputs], output)\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "5e7b843f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.keras.utils.plot_model(model, to_file='dense_plot.png', show_shapes=True, show_layer_names=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "8fb33d93",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(\n",
    "    loss = keras.losses.BinaryCrossentropy(), \n",
    "    optimizer = \"adam\",\n",
    "    metrics = ['acc']\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "008d0bd2",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "71/71 [==============================] - 17s 229ms/step - loss: 0.8151 - acc: 0.5625 - val_loss: 0.6426 - val_acc: 0.5894\n",
      "Epoch 2/20\n",
      "71/71 [==============================] - 16s 224ms/step - loss: 0.6667 - acc: 0.5966 - val_loss: 0.6045 - val_acc: 0.6755\n",
      "Epoch 3/20\n",
      "71/71 [==============================] - 16s 223ms/step - loss: 0.6156 - acc: 0.6534 - val_loss: 0.5323 - val_acc: 0.7285\n",
      "Epoch 4/20\n",
      "71/71 [==============================] - 16s 221ms/step - loss: 0.4993 - acc: 0.7670 - val_loss: 0.4312 - val_acc: 0.8278\n",
      "Epoch 5/20\n",
      "71/71 [==============================] - 16s 223ms/step - loss: 0.4861 - acc: 0.7841 - val_loss: 0.5425 - val_acc: 0.7483\n",
      "Epoch 6/20\n",
      "71/71 [==============================] - 16s 223ms/step - loss: 0.3991 - acc: 0.8125 - val_loss: 0.4359 - val_acc: 0.8212\n",
      "Epoch 7/20\n",
      "71/71 [==============================] - 16s 223ms/step - loss: 0.2874 - acc: 0.8864 - val_loss: 0.4911 - val_acc: 0.8079\n",
      "Epoch 8/20\n",
      "71/71 [==============================] - 16s 223ms/step - loss: 0.1894 - acc: 0.9489 - val_loss: 0.4566 - val_acc: 0.8344\n",
      "Epoch 9/20\n",
      "71/71 [==============================] - 16s 222ms/step - loss: 0.1036 - acc: 0.9744 - val_loss: 0.5550 - val_acc: 0.8079\n",
      "Epoch 10/20\n",
      "71/71 [==============================] - 16s 222ms/step - loss: 0.0897 - acc: 0.9631 - val_loss: 0.5494 - val_acc: 0.8742\n",
      "Epoch 11/20\n",
      "71/71 [==============================] - 16s 223ms/step - loss: 0.0762 - acc: 0.9659 - val_loss: 0.6126 - val_acc: 0.8146\n",
      "Epoch 12/20\n",
      "71/71 [==============================] - 16s 222ms/step - loss: 0.1346 - acc: 0.9545 - val_loss: 0.8376 - val_acc: 0.8013\n",
      "Epoch 13/20\n",
      "71/71 [==============================] - 16s 223ms/step - loss: 0.0789 - acc: 0.9744 - val_loss: 0.5270 - val_acc: 0.8146\n",
      "Epoch 14/20\n",
      "71/71 [==============================] - 16s 223ms/step - loss: 0.0460 - acc: 0.9858 - val_loss: 0.4637 - val_acc: 0.8940\n",
      "Epoch 15/20\n",
      "71/71 [==============================] - 16s 222ms/step - loss: 0.0326 - acc: 0.9858 - val_loss: 0.8294 - val_acc: 0.7815\n",
      "Epoch 16/20\n",
      "71/71 [==============================] - 16s 223ms/step - loss: 0.0309 - acc: 0.9886 - val_loss: 0.7629 - val_acc: 0.8411\n",
      "Epoch 17/20\n",
      "71/71 [==============================] - 16s 221ms/step - loss: 0.0107 - acc: 1.0000 - val_loss: 0.7001 - val_acc: 0.8543\n",
      "Epoch 18/20\n",
      "71/71 [==============================] - 16s 224ms/step - loss: 0.0058 - acc: 1.0000 - val_loss: 0.8859 - val_acc: 0.8477\n",
      "Epoch 19/20\n",
      "71/71 [==============================] - 16s 223ms/step - loss: 0.0157 - acc: 0.9915 - val_loss: 0.8888 - val_acc: 0.8411\n",
      "Epoch 20/20\n",
      "71/71 [==============================] - 16s 224ms/step - loss: 0.0074 - acc: 1.0000 - val_loss: 0.8373 - val_acc: 0.8411\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "callbacks = [\n",
    "    tf.keras.callbacks.ModelCheckpoint(\n",
    "        filepath='wv42model.{epoch:02d}-{val_loss:.2f}.h5',\n",
    "        save_best_only = False\n",
    "    ),\n",
    "]\n",
    "\n",
    "history = model.fit([X_train_spatial, X_train_feat], y_train_spatial, \n",
    "          batch_size = 5, \n",
    "          epochs = 20, \n",
    "          validation_data = ([X_test_spatial, X_test_feat], y_test_spatial), \n",
    "          callbacks = callbacks\n",
    "          )\n",
    "\n",
    "plt.plot(history.history['acc'])\n",
    "plt.plot(history.history['val_acc'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], \n",
    "           loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "5799ba22",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['acc'])\n",
    "plt.plot(history.history['val_acc'])\n",
    "plt.title('wv42 model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], \n",
    "           loc='upper left')\n",
    "plt.show()\n",
    "\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('wv42 model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], \n",
    "           loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "6ad92b8c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.011164051678609128,\n",
       " 0.011084591900860821,\n",
       " 0.010960828129646709,\n",
       " 0.01089244418674045,\n",
       " 0.010406331270319934,\n",
       " 0.011036432001979917,\n",
       " 0.011036766899956597,\n",
       " 0.011039048392422732,\n",
       " 0.011374362375183524,\n",
       " 0.01118948691829896,\n",
       " 0.01069765653466328,\n",
       " 0.010831117302956078,\n",
       " 0.011185661755471563,\n",
       " 0.011165370339392307,\n",
       " 0.011119808516221118,\n",
       " 0.011368969994480881,\n",
       " 0.011347424019184295,\n",
       " 0.010742999889232495,\n",
       " 0.010839186905179331,\n",
       " 0.011319091127079047,\n",
       " 0.010878827182680817,\n",
       " 0.010717075384529857,\n",
       " 0.011071768971300583,\n",
       " 0.010968674030487102,\n",
       " 0.010947077035576881,\n",
       " 0.011155854527351787,\n",
       " 0.011073338805566272,\n",
       " 0.01112151963557072,\n",
       " 0.010983366371018733,\n",
       " 0.011049134577579786,\n",
       " 0.011239694142701368,\n",
       " 0.011190523008914315,\n",
       " 0.01062496797538098,\n",
       " 0.010838404604436929,\n",
       " 0.01125475146969977,\n",
       " 0.010396351048975815,\n",
       " 0.010875077895176263,\n",
       " 0.010814664131656432,\n",
       " 0.01100458791390041,\n",
       " 0.010934339138706034,\n",
       " 0.011236779483748072,\n",
       " 0.010962206313312463,\n",
       " 0.011086949268649799,\n",
       " 0.011063624147852098,\n",
       " 0.011190422277882266,\n",
       " 0.011064999061029798,\n",
       " 0.010723325941297743,\n",
       " 0.011183420817057291,\n",
       " 0.01078097548831474,\n",
       " 0.011146864609790273,\n",
       " 0.01078895744145816,\n",
       " 0.01094624960209934,\n",
       " 0.011000779757610891,\n",
       " 0.011118539566856354,\n",
       " 0.010467479585456586,\n",
       " 0.010940490272637094,\n",
       " 0.010874719449685597,\n",
       " 0.011015027311767242,\n",
       " 0.010838073631045914,\n",
       " 0.011100517869486239,\n",
       " 0.010782821351772146,\n",
       " 0.010631503064609523,\n",
       " 0.010428707295484504,\n",
       " 0.011135744950408307,\n",
       " 0.011459210609048809,\n",
       " 0.011176526628240473,\n",
       " 0.011098989897467635,\n",
       " 0.010992453095026632,\n",
       " 0.010459731145160188,\n",
       " 0.011273947926378708,\n",
       " 0.01085496439364712,\n",
       " 0.011128661073284385,\n",
       " 0.010913750108213908,\n",
       " 0.010989373603475436,\n",
       " 0.010593322732975127,\n",
       " 0.010911653725371603,\n",
       " 0.01067191648548686,\n",
       " 0.010561710195449808,\n",
       " 0.010907470117053541,\n",
       " 0.010968586381407268,\n",
       " 0.010855932458110961,\n",
       " 0.010909136103668003,\n",
       " 0.01107065962175283,\n",
       " 0.011263199794439622,\n",
       " 0.011138700163413468,\n",
       " 0.011124566422241378,\n",
       " 0.010663442650940193,\n",
       " 0.01101208779710474,\n",
       " 0.01088439420744552,\n",
       " 0.010894356768154149,\n",
       " 0.011194396574964903,\n",
       " 0.01120258064427003,\n",
       " 0.011001590838648163,\n",
       " 0.010888124526119362,\n",
       " 0.010662755848448953,\n",
       " 0.01101292373385122,\n",
       " 0.010920612900345414,\n",
       " 0.011256334385917674,\n",
       " 0.011297347614304029,\n",
       " 0.010869255772343388,\n",
       " 0.010792512461972334,\n",
       " 0.010983087725436573,\n",
       " 0.011015704956891933,\n",
       " 0.010933558146158854,\n",
       " 0.010978561369970502,\n",
       " 0.011366536751369062,\n",
       " 0.01143942677271546,\n",
       " 0.011064605294268155,\n",
       " 0.010887927642738542,\n",
       " 0.01128582633898909,\n",
       " 0.011171388037410784,\n",
       " 0.011569932328003097,\n",
       " 0.011065311719687713,\n",
       " 0.010758345673277218,\n",
       " 0.011207941628287359,\n",
       " 0.010940825170613774,\n",
       " 0.010456304981875322,\n",
       " 0.01100136321267964,\n",
       " 0.01087411702578614,\n",
       " 0.010662262658850483,\n",
       " 0.0103759216183007,\n",
       " 0.010928635407540696,\n",
       " 0.010515623785994835,\n",
       " 0.010787152786150228,\n",
       " 0.010951624322166495,\n",
       " 0.010727782308319469,\n",
       " 0.011070933034554104,\n",
       " 0.010690148802287622,\n",
       " 0.011409172141829963,\n",
       " 0.010975917507428037,\n",
       " 0.011038600989657012,\n",
       " 0.011235067056203248,\n",
       " 0.011160598043224612,\n",
       " 0.010872486360442656,\n",
       " 0.01075772624283988,\n",
       " 0.01103647255603178,\n",
       " 0.011659306917334453,\n",
       " 0.010527888770293143,\n",
       " 0.010891152998056921,\n",
       " 0.010904883161003206,\n",
       " 0.011032933887957872,\n",
       " 0.010907558420230987,\n",
       " 0.01142659468578867,\n",
       " 0.0109397145128708]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conv1_weights = model.layers[1].get_weights()[0]\n",
    "\n",
    "conv1_weights_abs = np.absolute(conv1_weights)\n",
    "\n",
    "list_avgs = []\n",
    "\n",
    "for i in range(conv1_weights.shape[2]):\n",
    "\n",
    "    hold_weights = conv1_weights_abs[:, :, i, :]\n",
    "    hold_sum = hold_weights.sum()\n",
    "    hold_avg = hold_sum / 729\n",
    "    list_avgs.append(hold_avg)\n",
    "    \n",
    "list_avgs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "ce581c6d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(8, 0.012398821826824925),\n",
       " (1, 0.01321332117821127),\n",
       " (0, 0.013612129724238309),\n",
       " (2, 0.014373614464277103),\n",
       " (3, 0.01502382019419729),\n",
       " (5, 0.015465269363466113),\n",
       " (4, 0.015606477263860087),\n",
       " (7, 0.01616925238254793),\n",
       " (6, 0.01617244830347383),\n",
       " (10, 0.017187160392520523),\n",
       " (9, 0.017216333145957916),\n",
       " (36, 0.01760433076667524),\n",
       " (17, 0.01765285696021814),\n",
       " (32, 0.01766622933177137),\n",
       " (12, 0.017883740334844394),\n",
       " (30, 0.018115822016293456),\n",
       " (31, 0.018275063387814535),\n",
       " (27, 0.01847057839643481),\n",
       " (40, 0.01874487292128826),\n",
       " (44, 0.018813126698769986),\n",
       " (35, 0.018834090527193047),\n",
       " (28, 0.01889277878121584),\n",
       " (13, 0.018979916356718588),\n",
       " (14, 0.018985644943593134),\n",
       " (11, 0.019091269933964816),\n",
       " (39, 0.01936148812250836),\n",
       " (37, 0.019454427543818053),\n",
       " (29, 0.01966071684828184),\n",
       " (111, 0.019780709746770243),\n",
       " (38, 0.019850860407323014),\n",
       " (18, 0.020317404685524103),\n",
       " (76, 0.020412947415324397),\n",
       " (26, 0.020443122246301387),\n",
       " (54, 0.02051339587720497),\n",
       " (50, 0.02064899816107521),\n",
       " (65, 0.020724292645238555),\n",
       " (108, 0.02073070934279956),\n",
       " (20, 0.02074224108367627),\n",
       " (113, 0.020883914701569063),\n",
       " (49, 0.02089903743832854),\n",
       " (41, 0.02092889699425717),\n",
       " (142, 0.020958230655706796),\n",
       " (23, 0.021125916427382062),\n",
       " (112, 0.02113711065389165),\n",
       " (24, 0.02117721076201346),\n",
       " (75, 0.021198673012815875),\n",
       " (15, 0.021234856384444792),\n",
       " (116, 0.02129351978275838),\n",
       " (117, 0.021423352925702212),\n",
       " (141, 0.021425629185387464),\n",
       " (91, 0.021437347998180836),\n",
       " (131, 0.021489538446538895),\n",
       " (63, 0.021656384356882018),\n",
       " (19, 0.02169510039445958),\n",
       " (64, 0.021709070611228995),\n",
       " (90, 0.02171877741977184),\n",
       " (96, 0.021760310001661093),\n",
       " (134, 0.02176682088927804),\n",
       " (132, 0.02182563603823731),\n",
       " (88, 0.02182900071634677),\n",
       " (56, 0.021842366546923894),\n",
       " (59, 0.021879957536611046),\n",
       " (119, 0.021891731293603715),\n",
       " (127, 0.021963276490262508),\n",
       " (98, 0.021975403459964956),\n",
       " (82, 0.02198169849537037),\n",
       " (55, 0.022040248080699695),\n",
       " (61, 0.022073703865291977),\n",
       " (94, 0.02212023636931745),\n",
       " (58, 0.022187807268893606),\n",
       " (45, 0.022192338857140562),\n",
       " (74, 0.02220710576479981),\n",
       " (140, 0.02227624388225121),\n",
       " (66, 0.022287036492827823),\n",
       " (71, 0.022313286738140592),\n",
       " (73, 0.022321491738569264),\n",
       " (139, 0.022322370845758048),\n",
       " (83, 0.02232551313067987),\n",
       " (125, 0.022353796311366705),\n",
       " (72, 0.022409187913431553),\n",
       " (16, 0.022417746126570003),\n",
       " (42, 0.02245200514302823),\n",
       " (97, 0.022457243156694746),\n",
       " (110, 0.022486300788952654),\n",
       " (100, 0.02250903722190072),\n",
       " (51, 0.022513770272211773),\n",
       " (124, 0.02256703213245617),\n",
       " (89, 0.022609336549526054),\n",
       " (60, 0.022692527300045815),\n",
       " (126, 0.022698892977993185),\n",
       " (77, 0.022715852420843514),\n",
       " (21, 0.02272766019091194),\n",
       " (104, 0.022739394702047968),\n",
       " (53, 0.022749499201938122),\n",
       " (109, 0.02276847326542942),\n",
       " (107, 0.02279995629162782),\n",
       " (62, 0.022808412465539),\n",
       " (81, 0.022883400832020534),\n",
       " (47, 0.02292163676195184),\n",
       " (129, 0.023029031740458086),\n",
       " (123, 0.023064860591182002),\n",
       " (86, 0.023117667199489347),\n",
       " (128, 0.023129017101230283),\n",
       " (84, 0.023129378163111388),\n",
       " (80, 0.02315868696885181),\n",
       " (101, 0.023221114044817387),\n",
       " (22, 0.023235860021353093),\n",
       " (52, 0.02325720976736647),\n",
       " (68, 0.02326677267443496),\n",
       " (67, 0.02328529410237965),\n",
       " (46, 0.023328475010247877),\n",
       " (118, 0.023469639740199544),\n",
       " (85, 0.02347421580708403),\n",
       " (133, 0.023494644583661534),\n",
       " (93, 0.02350219810286994),\n",
       " (137, 0.023561556152846098),\n",
       " (121, 0.023567775312928668),\n",
       " (95, 0.02358652959962277),\n",
       " (92, 0.023696355204850396),\n",
       " (122, 0.023700415842817645),\n",
       " (48, 0.024007093432181822),\n",
       " (99, 0.024050904890146767),\n",
       " (79, 0.024072620930822134),\n",
       " (25, 0.024102937046883037),\n",
       " (103, 0.024211145722817),\n",
       " (57, 0.024300916695300444),\n",
       " (115, 0.024357360086323304),\n",
       " (114, 0.024378186554248112),\n",
       " (34, 0.024403298669718257),\n",
       " (43, 0.024464082652485748),\n",
       " (70, 0.0245369883721748),\n",
       " (33, 0.024579601523316938),\n",
       " (136, 0.02481394376610859),\n",
       " (105, 0.024844594780146178),\n",
       " (135, 0.024887650101273147),\n",
       " (106, 0.02490512497304071),\n",
       " (102, 0.025040551958751284),\n",
       " (69, 0.02508392647951228),\n",
       " (143, 0.02527188011975282),\n",
       " (138, 0.02538870195302453),\n",
       " (120, 0.025646727762104554),\n",
       " (78, 0.025741163758747534),\n",
       " (87, 0.02596789492174104),\n",
       " (130, 0.027206551717959642)]"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list_avgs_sort = sorted(enumerate(list_avgs), key = lambda x: x[1])\n",
    "\n",
    "list_avgs_sort"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "4af23ff3",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'X_test_spatial' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-7-3146c8afd8b9>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave_weights\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfilename\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mepoch\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m14\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m \u001b[0my_pred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mX_test_spatial\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX_test_feat\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      6\u001b[0m \u001b[0my_pred_round\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_pred\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'X_test_spatial' is not defined"
     ]
    }
   ],
   "source": [
    "filename = 'wv42_model_weights'\n",
    "\n",
    "model.save_weights(filename.format(epoch = 14))\n",
    "\n",
    "y_pred = model.predict([X_test_spatial, X_test_feat], )\n",
    "y_pred_round = np.rint(y_pred)\n",
    "\n",
    "y_pred_round = y_pred_round.reshape(y_pred_round.shape[0])\n",
    "\n",
    "print(y_pred_round)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "1ca7778a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "precision:  0.9402985074626866\n",
      "recall:  0.84\n",
      "f1 score:  0.8873239436619718\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import precision_score, recall_score, f1_score\n",
    "\n",
    "print(\"precision: \", precision_score(y_test_spatial, y_pred_round))\n",
    "print(\"recall: \", recall_score(y_test_spatial, y_pred_round))\n",
    "print(\"f1 score: \", f1_score(y_test_spatial, y_pred_round))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "4b3e197e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEGCAYAAABo25JHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAAAiWUlEQVR4nO3deXxV1bn/8c8js0O0F9B6GQyTjEKQiKK/WKm1xQEp1QaoaFFbK4ri1FdpqVqx11vBWvXWiqiIQ4uoV2hE1B8VKBRQDDKDlAioQVBEfoDVKODz+2PvpIeMJyb7nCT7+369zos9rL3Ps3NCnrPW2nstc3dERCS+Dkt3ACIikl5KBCIiMadEICISc0oEIiIxp0QgIhJzjdMdQHW1atXKMzMz0x2GiEi9snz58o/dvXV5++pdIsjMzCQ/Pz/dYYiI1Ctm9m5F+9Q0JCISc0oEIiIxp0QgIhJzSgQiIjGnRCAiEnORJQIzm2pmH5nZ2gr2m5k9YGYFZrbazE6OKhYREalYlDWCacCgSvafC3QJX1cBD0UYi4iIVCCy5wjcfaGZZVZSZAjwpAfjYL9uZseY2fHuvj2qmEREiv3ljff468pt6Q6jWnr8Zwa3D+5Z6+dNZx9BG+D9hPXCcFsZZnaVmeWbWf7OnTtTEpyINGx/XbmN9dv3pjuMOqFePFns7lOAKQDZ2dmaSUdEakWP4zOY8bMB6Q4j7dKZCLYB7RLW24bbRKQBqOtNL+u376XH8RnpDqNOSGfTUB5wWXj30GnAHvUPiDQcdb3ppcfxGQzJKrc1OnYiqxGY2XTgLKCVmRUCtwNNANx9MjAHOA8oAD4DLo8qFhFJDzW91A9R3jU0oor9Dlwb1fuLSLSqavpR00v9oSeLReRrqarpR00v9Ue9uGtIpK6r6x2jUSj+xq+mn/pPNQKRWlDXO0ajoG/8DYdqBCK1RN+Opb5SjUBEJOaUCEREYk5NQ1In1bfOV90qKfWZagRSJ9W3zld1nEp9phqB1FnqfBVJDSWCmKrrTS9qahFJHTUNxVRdb3pRU4tI6qhGEGNqehERUCKoM1LdVKOmFxEppqahOiLVTTVqehGRYqoRpEnpGoAG8BKRdFGNIE1K1wD0DV1E0kU1gjRSDUBE6gIlghSpqClIRCTd1DSUImoKEpG6SjWCFFJTkIjURUoEX1N17/tXU5CI1FVqGvqaqnvfv5qCRKSuUo0gSbrvX0QaKtUIkqTOXhFpqFQjqAbVAESkIVKNQEQk5pQIRERiTolARCTmlAhERGJOiUBEJOZie9eQngwWEQlEWiMws0FmttHMCsxsXDn725vZfDNbYWarzey8KONJpCeDRUQCkdUIzKwR8CBwDlAIvGlmee6+PqHYr4Fn3f0hM+sBzAEyo4qpND0XICISbY2gP1Dg7pvd/UvgGWBIqTIOFLe3HA18EGE8IiJSjigTQRvg/YT1wnBbot8AI82skKA2cF15JzKzq8ws38zyd+7cGUWsIiKxle67hkYA09y9LXAe8JSZlYnJ3ae4e7a7Z7du3TrlQYqINGRRJoJtQLuE9bbhtkRXAs8CuPtSoDnQKsKYRESklCgTwZtAFzPrYGZNgeFAXqky7wFnA5hZd4JEoLYfEZEUiiwRuPsBYAzwKrCB4O6gdWY2wcwuDIvdDPzUzFYB04FR7u5RxSQiImVF+kCZu88h6ARO3HZbwvJ64IwoYxARkcqlu7NYRETSTIlARCTmlAhERGJOiUBEJOaUCEREYk6JQEQk5pQIRERiTolARCTmlAhERGIuNlNVlp6aUlNPiogEkq4RmNnhUQYStdJTU2rqSRGRQJU1AjM7HXgUOBJob2Z9gJ+5+zVRB1fbNDWliEhZydQI/gB8D9gF4O6rgDOjDEpERFInqaYhd3+/1KaDEcQiIiJpkExn8fth85CbWRNgLMH8AiIi0gAkUyO4GriWYOL5bUAWUO/6B0REpHzJ1Ai6uvsliRvM7AxgcTQhiYhIKiVTI/ifJLeJiEg9VGGNwMwGAKcDrc3spoRdGUCjqAMTEZHUqKxpqCnBswONgaMStu8FLo4yKBERSZ0KE4G7/x34u5lNc/d3UxiTiIikUDKdxZ+Z2SSgJ9C8eKO7fzuyqEREJGWS6Sz+M/A20AG4A9gKvBlhTCIikkLJJIKW7v4YsN/d/+7uVwCqDYiINBDJNA3tD//dbmbnAx8A/xFdSCIikkrJJILfmtnRwM0Ezw9kADdEGZSIiKROlYnA3WeHi3uAgVDyZLGIiDQAlT1Q1gjIJRhj6BV3X2tmFwC/AloAfVMTooiIRKmyGsFjQDtgGfCAmX0AZAPj3H1WCmITEZEUqCwRZAO93f0rM2sO7AA6ufuu1IQmIiKpUNnto1+6+1cA7l4EbK5uEjCzQWa20cwKzGxcBWVyzWy9ma0zs79U5/wiIlJzldUIupnZ6nDZgE7hugHu7r0rO3HYx/AgcA5QCLxpZnnuvj6hTBfgl8AZ7r7bzI6twbWIiMjXUFki6F7Dc/cHCtx9M4CZPQMMAdYnlPkp8KC77wZw949q+J4iIlJNlQ06V9OB5toAiXMdFwKnlipzIoCZLSYY2vo37v5K6ROZ2VXAVQDt27evYVgiIpIoqcnrI9QY6AKcBYwAHjGzY0oXcvcp7p7t7tmtW7dObYQiIg1clIlgG8Htp8XahtsSFQJ57r7f3bcA/yRIDCIikiJJJQIza2FmXat57jeBLmbWwcyaAsOBvFJlZhHUBjCzVgRNRZur+T4iIlIDVSYCMxsMrAReCdezzKz0H/Qy3P0AMAZ4FdgAPOvu68xsgpldGBZ7FdhlZuuB+cDP9ZyCiEhqJTPo3G8I7gBaAODuK82sQzInd/c5wJxS225LWHbgpvAlIiJpkEzT0H5331Nqm0cRjIiIpF4yNYJ1ZvYjoFH4ANj1wJJowxIRkVRJpkZwHcF8xV8AfyEYjvqGCGMSEZEUSqZG0M3dxwPjow5GRERSL5kawe/NbIOZ3WlmvSKPSEREUqrKRODuAwlmJtsJPGxma8zs15FHJiIiKZHUA2XuvsPdHwCuJnim4LbKjxARkfoimQfKupvZb8xsDcHk9UsIhosQEZEGIJnO4qnADOB77v5BxPGIiEiKVZkI3H1AKgIREZH0qDARmNmz7p4bNgklPkmc1AxlIiJSP1RWIxgb/ntBKgIREZH0qLCz2N23h4vXuPu7iS/gmtSEJyIiUUvm9tFzytl2bm0HIiIi6VFZH8Fogm/+Hc1sdcKuo4DFUQcmIiKpUVkfwV+Al4H/BsYlbN/n7p9EGpWIiKRMZYnA3X2rmV1beoeZ/YeSgYhIw1BVjeACYDnB7aOWsM+BjhHGJSIiKVJhInD3C8J/k5qWUkRE6qdkxho6w8yOCJdHmtm9ZtY++tBERCQVkrl99CHgMzPrA9wMvAM8FWlUIiKSMskkggPu7sAQ4I/u/iDBLaQiItIAJDP66D4z+yVwKZBjZocBTaINS0REUiWZGsEwgonrr3D3HQRzEUyKNCoREUmZZKaq3AH8GTjazC4Aitz9ycgjExGRlEjmrqFcYBnwQyAXeMPMLo46MBERSY1k+gjGA6e4+0cAZtYa+BvwfJSBiYhIaiTTR3BYcRII7UryOBERqQeSqRG8YmavAtPD9WHAnOhCEhGRVEpmzuKfm9kPgP8Tbpri7jOjDUtERFKlsvkIugD3AJ2ANcAt7r4tVYGJiEhqVNbWPxWYDVxEMALp/1T35GY2yMw2mlmBmY2rpNxFZuZmll3d9xARkZqprGnoKHd/JFzeaGZvVefEZtYIeJBgqstC4E0zy3P39aXKHQWMBd6ozvlFRKR2VJYImptZX/49D0GLxHV3ryox9AcK3H0zgJk9QzBe0fpS5e4E7gZ+Xs3YRUSkFlSWCLYD9yas70hYd+DbVZy7DfB+wnohcGpiATM7GWjn7i+ZWYWJwMyuAq4CaN9eI2CLiNSmyiamGRjlG4eD190LjKqqrLtPAaYAZGdne5RxiYjETZQPhm0D2iWstw23FTsK6AUsMLOtwGlAnjqMRURSK8pE8CbQxcw6mFlTYDiQV7zT3fe4eyt3z3T3TOB14EJ3z48wJhERKSWyRODuB4AxwKvABuBZd19nZhPM7MKo3ldERKqnyieLzcyAS4CO7j4hnK/4m+6+rKpj3X0OpYajcPfbKih7VlIRi4hIrUqmRvAnYAAwIlzfR/B8gIiINADJDDp3qrufbGYrANx9d9jmLyIiDUAyNYL94VPCDiXzEXwVaVQiIpIyySSCB4CZwLFm9l/AP4C7Io1KRERSJplhqP9sZsuBswmGl/i+u2+IPDIREUmJZO4aag98BryYuM3d34syMBERSY1kOotfIugfMKA50AHYCPSMMC4REUmRZJqGTkpcDweKuyayiEREJKWq/WRxOPz0qVUWFBGReiGZPoKbElYPA04GPogsIhERSalk+giOSlg+QNBn8L/RhCMiIqlWaSIIHyQ7yt1vSVE8IiKSYhX2EZhZY3c/CJyRwnhERCTFKqsRLCPoD1hpZnnAc8C/ine6+wsRxyYiIimQTB9Bc2AXwRzFxc8TOKBEICLSAFSWCI4N7xhay78TQDHNGywi0kBUlggaAUdyaAIopkQgItJAVJYItrv7hJRFIiIiaVHZk8Xl1QRERKSBqSwRnJ2yKEREJG0qTATu/kkqAxERkfSo9qBzIiLSsCgRiIjEnBKBiEjMKRGIiMScEoGISMwpEYiIxJwSgYhIzCkRiIjEnBKBiEjMRZoIzGyQmW00swIzG1fO/pvMbL2ZrTaz18zshCjjERGRsiJLBOF8xw8C5wI9gBFm1qNUsRVAtrv3Bp4HJkYVj4iIlC/KGkF/oMDdN7v7l8AzwJDEAu4+390/C1dfB9pGGI+IiJQjykTQBng/Yb0w3FaRK4GXy9thZleZWb6Z5e/cubMWQxQRkTrRWWxmI4FsYFJ5+919irtnu3t269atUxuciEgDl8zk9V/XNqBdwnrbcNshzOw7wHjgW+7+RYTxiIhIOaKsEbwJdDGzDmbWFBgO5CUWMLO+wMPAhe7+UYSxiIhIBSJLBO5+ABgDvApsAJ5193VmNsHMLgyLTQKOBJ4zs5VmllfB6UREJCJRNg3h7nOAOaW23Zaw/J0o319ERKpWJzqLRUQkfZQIRERiTolARCTmlAhERGJOiUBEJOaUCEREYk6JQEQk5pQIRERiTolARCTmlAhERGJOiUBEJOaUCEREYk6JQEQk5pQIRERiTolARCTmlAhERGJOiUBEJOaUCEREYk6JQEQk5pQIRERiTolARCTmGqc7AJFU2b9/P4WFhRQVFaU7FJHING/enLZt29KkSZOkj1EikNgoLCzkqKOOIjMzEzNLdzgitc7d2bVrF4WFhXTo0CHp49Q0JLFRVFREy5YtlQSkwTIzWrZsWe1arxKBxIqSgDR0X+d3XIlARCTmlAhE6omDBw/St29fLrjggpJtl1xyCV27dqVXr15cccUV7N+/v8xxCxYs4OijjyYrK4tu3bpxyy23HLJ/1qxZ9O7dm+7du3PSSScxa9asQ/bfc889dOvWjaysLE455RSefPLJSK6vJu67775D4jpw4ACtW7dm3Lhxh5TLzMzk448/LllfsGDBIT/Pl19+mezsbHr06EHfvn25+eabaxzb8uXLOemkk+jcuTPXX3897l6mzO7duxk6dCi9e/emf//+rF279pD95X32w4cPZ9OmTTWOD5QIROqN+++/n+7dux+y7ZJLLuHtt99mzZo1fP755zz66KPlHpuTk8PKlStZsWIFs2fPZvHixQCsWrWKW265hb/+9a9s2LCBvLw8brnlFlavXg3A5MmTmTt3LsuWLWPlypW89tpr5f4hq4mDBw/W6PgDBw4wdepUfvSjH5Vsmzt3LieeeCLPPfdc0vGuXbuWMWPG8PTTT7N+/Xry8/Pp3LlzjWIDGD16NI888gibNm1i06ZNvPLKK2XK3HXXXWRlZbF69WqefPJJxo4de8j+8j770aNHM3HixBrHB7prSGLqjhfXsf6DvbV6zh7/mcHtg3tWuH/SpEk0a9aM66+/nhtvvJFVq1Yxb9485s2bx2OPPUZOTg7vvPMOkyZNAmDatGnk5+fzxz/+kcLCQl566SXGjx/PvffeW3LO8847r2S5f//+FBYWVhpjixYtyMrKYtu2bUDwbf9Xv/pVyR0mHTp04Je//CWTJk3iqaee4q677mLBggVkZGQAkJGRwY9//OMy5y0oKODqq69m586dNGrUiOeee47333+fe+65h9mzZwMwZswYsrOzGTVqFJmZmQwbNoy5c+eSm5vLCy+8wLJlywDYunUrgwcPZs2aNSxfvpybbrqJTz/9lFatWjFt2jSOP/74Q9573rx5nHzyyTRu/O8/Z9OnT2fs2LE89NBDLF26lNNPP73SnwvAxIkTGT9+PN26dQOgUaNGjB49usrjKrN9+3b27t3LaaedBsBll13GrFmzOPfccw8pt379+pLaS7du3di6dSsffvghxx13XIWffU5ODqNGjeLAgQOHXPvXoRqBSIrk5OSwaNEiAPLz8/n000/Zv38/ixYt4swzz+Siiy5i5syZJeVnzJjB8OHDAbjhhhuYOHEihx1W/n/Z/fv389RTTzFo0KBKY9i9ezebNm3izDPPBGDdunX069fvkDLZ2dmsW7eOvXv3sm/fPjp27FjltV1yySVce+21rFq1iiVLlpT5Y12eli1b8tZbbzFu3Di+/PJLtmzZUnLdw4YNY//+/Vx33XU8//zzLF++nCuuuILx48eXOc/ixYsPuYaioiL+9re/MXjwYEaMGMH06dOrjAWCGkHpn0V55s+fT1ZWVplXeclm27ZttG3btmS9bdu2JUk4UZ8+fXjhhRcAWLZsGe+++25JUq/osz/ssMPo3Lkzq1atSur6KqMagcRSZd/co9KvXz+WL1/O3r17adasGSeffDL5+fksWrSIBx54gNatW9OxY0def/11unTpwttvv80ZZ5zB7NmzOfbYY+nXrx8LFiwo99zXXHMNZ555Jjk5OeXuX7RoEX369GHTpk3ccMMNfPOb36y169q3bx/btm1j6NChQPBAUzKGDRtWspybm8uMGTMYN24cM2bMYMaMGWzcuJG1a9dyzjnnAEETUnkJZvv27Yc0m8yePZuBAwfSokULLrroIu68807uu+8+GjVqVO4dNdW9y2bgwIGsXLmyWsdUZdy4cYwdO5asrCxOOukk+vbtS6NGjar87I899lg++OCDpBJYZSJNBGY2CLgfaAQ86u6/K7W/GfAk0A/YBQxz961RxiSSLk2aNKFDhw5MmzaN008/nd69ezN//nwKCgpK/pANHz6cZ599lm7dujF06FDMjMWLF5OXl8ecOXMoKipi7969jBw5kqeffhqAO+64g507d/Lwww9X+N45OTnMnj2bLVu2cNppp5Gbm0tWVhY9evRg+fLl9OnTp6Ts8uXL6dmzJxkZGRx55JFs3rw5qVpBaY0bN+arr74qWS99b/sRRxxRsjxs2DB++MMf8oMf/AAzo0uXLqxZs4aePXuydOnSSt+nRYsWh5x7+vTp/OMf/yAzMxOAXbt2MW/ePM455xxatmzJ7t27adWqFQCffPJJyXLPnj3L/CzKM3/+fG688cYy2w8//HCWLFlyyLY2bdoc0lxXWFhImzZtyhybkZHB448/DgQPhXXo0IGOHTsyY8aMSj/7oqIiWrRoUWm8SXH3SF4Ef/zfAToCTYFVQI9SZa4BJofLw4EZVZ23X79+/nXkTl7iuZOXfK1jpWFYv359ukPw22+/3du1a+dz5871HTt2eLt27fz73/9+yf5PPvnEO3bs6GeddZa/8cYbZY6fP3++n3/++SXrjzzyiA8YMMA/++yzCt+z9DH33nuvDx8+3N3dV6xY4Z07d/YtW7a4u/uWLVu8U6dOvmLFCnd3f/DBB33QoEG+Z88ed3fft2+fP/HEE2Xe49RTT/WZM2e6u3tRUZH/61//8vfee89POOEELyoq8t27d3tmZqY//vjj7u5+wgkn+M6dOw85R3Z2to8cOdLvvvtud3f/4osvvFOnTr5kSfD/9ssvv/S1a9eWee+HHnrIx48f7+7ue/bs8datW3tRUVHJ/qlTp/rll1/u7u4333yz33rrre7ufuDAAR86dGjJ9axatco7derkGzdudHf3gwcP+kMPPVThzzVZp5xyii9dutS/+uorHzRokL/00ktlyuzevdu/+OILd3efMmWKX3rppWXKlP4c3d179erl27dvL1O2vN91IN8r+LsaZR9Bf6DA3Te7+5fAM8CQUmWGAE+Ey88DZ5ue+JEGLCcnh+3btzNgwACOO+44mjdvfkhzzje+8Q26d+/Ou+++S//+/as839VXX82HH37IgAEDyMrKYsKECUkds3DhQrZu3UpWVhZ33303gwcPplu3bgwePJiJEyeSlZUFBHemDBw4kFNOOYVevXqRk5NTbj/FU089xQMPPEDv3r05/fTT2bFjB+3atSM3N5devXqRm5tL3759K41r2LBhPP300+Tm5gLQtGlTnn/+eX7xi1/Qp08fsrKyynzjBjj33HNZuHAhADNnzuTb3/42zZo1K9k/ZMgQXnzxRb744gtuvfVWCgoK6NOnD3379qVz586MHDkSgN69e3PfffcxYsQIunfvTq9evdi8eXOVP8+q/OlPf+InP/kJnTt3plOnTiUdxZMnT2by5MkAbNiwgV69etG1a1defvll7r///irP++GHH9KiRYtaaeYzr+VbwUpObHYxMMjdfxKuXwqc6u5jEsqsDcsUhuvvhGU+LnWuq4CrANq3b9/v3XffrXY8d7y4DkhP27DUDRs2bChzC540DEOHDmXixIl06dIl3aGkzB/+8AcyMjK48sory+wr73fdzJa7e3Z556oXncXuPgWYApCdnf21MpcSgEjD9bvf/Y7t27fHKhEcc8wxXHrppbVyrigTwTagXcJ623BbeWUKzawxcDRBp7GISNK6du1K165d0x1GSl1++eW1dq4o+wjeBLqYWQcza0rQGZxXqkweUPx0ysXAPI+qrUoEav2pWJG65uv8jkeWCNz9ADAGeBXYADzr7uvMbIKZXRgWewxoaWYFwE3AuPLPJlJzzZs3Z9euXUoG0mB5OB9Bss9yFIusszgq2dnZnp+fn+4wpB7SDGUSBxXNUFbvO4tFakPxA10iciiNNSQiEnNKBCIiMadEICISc/Wus9jMdgLVf7Q40Ar4uMpSDYuuOR50zfFQk2s+wd1bl7ej3iWCmjCz/Ip6zRsqXXM86JrjIaprVtOQiEjMKRGIiMRc3BLBlHQHkAa65njQNcdDJNccqz4CEREpK241AhERKUWJQEQk5hpkIjCzQWa20cwKzKzMiKZm1szMZoT73zCzzDSEWauSuOabzGy9ma02s9fM7IR0xFmbqrrmhHIXmZmbWb2/1TCZazaz3PCzXmdmf0l1jLUtid/t9mY238xWhL/f56UjztpiZlPN7KNwBsfy9puZPRD+PFab2ck1ftOKJjOury+gEfAO0BFoCqwCepQqcw0wOVweDsxId9wpuOaBwOHh8ug4XHNY7ihgIfA6kJ3uuFPwOXcBVgDfCNePTXfcKbjmKcDocLkHsDXdcdfwms8ETgbWVrD/POBlwIDTgDdq+p4NsUbQHyhw983u/iXwDDCkVJkhwBPh8vPA2WZmKYyxtlV5ze4+390/C1dfJ5gxrj5L5nMGuBO4G2gIY08nc80/BR50990A7v5RimOsbclcswMZ4fLRwAcpjK/WuftC4JNKigwBnvTA68AxZnZ8Td6zISaCNsD7CeuF4bZyy3gwgc4eoGVKootGMtec6EqCbxT1WZXXHFaZ27n7S6kMLELJfM4nAiea2WIze93MBqUsumgkc82/AUaaWSEwB7guNaGlTXX/v1dJ8xHEjJmNBLKBb6U7liiZ2WHAvcCoNIeSao0JmofOIqj1LTSzk9z9/6UzqIiNAKa5++/NbADwlJn1cvev0h1YfdEQawTbgHYJ623DbeWWMbPGBNXJXSmJLhrJXDNm9h1gPHChu3+RotiiUtU1HwX0AhaY2VaCttS8et5hnMznXAjkuft+d98C/JMgMdRXyVzzlcCzAO6+FGhOMDhbQ5XU//fqaIiJ4E2gi5l1MLOmBJ3BeaXK5AE/DpcvBuZ52AtTT1V5zWbWF3iYIAnU93ZjqOKa3X2Pu7dy90x3zyToF7nQ3evzPKfJ/G7PIqgNYGatCJqKNqcwxtqWzDW/B5wNYGbdCRLBzpRGmVp5wGXh3UOnAXvcfXtNTtjgmobc/YCZjQFeJbjjYKq7rzOzCUC+u+cBjxFUHwsIOmWGpy/imkvymicBRwLPhf3i77n7hWkLuoaSvOYGJclrfhX4rpmtBw4CP3f3elvbTfKabwYeMbMbCTqOR9XnL3ZmNp0gmbcK+z1uB5oAuPtkgn6Q84AC4DPg8hq/Zz3+eYmISC1oiE1DIiJSDUoEIiIxp0QgIhJzSgQiIjGnRCAiEnNKBFInmdlBM1uZ8MqspOyntfB+08xsS/heb4VPqFb3HI+aWY9w+Vel9i2paYzheYp/LmvN7EUzO6aK8ln1fTROiZ5uH5U6ycw+dfcja7tsJeeYBsx29+fN7LvAPe7euwbnq3FMVZ3XzJ4A/unu/1VJ+VEEo66Oqe1YpOFQjUDqBTM7MpxH4S0zW2NmZUYaNbPjzWxhwjfmnHD7d81saXjsc2ZW1R/ohUDn8NibwnOtNbMbwm1HmNlLZrYq3D4s3L7AzLLN7HdAizCOP4f7Pg3/fcbMzk+IeZqZXWxmjcxskpm9GY4x/7MkfixLCQcbM7P+4TWuMLMlZtY1fBJ3AjAsjGVYGPtUM1sWli1vxFaJm3SPva2XXuW9CJ6KXRm+ZhI8BZ8R7mtF8FRlcY320/Dfm4Hx4XIjgvGGWhH8YT8i3P4L4LZy3m8acHG4/EPgDaAfsAY4guCp7HVAX+Ai4JGEY48O/11AOOdBcUwJZYpjHAo8ES43JRhFsgVwFfDrcHszIB/oUE6cnyZc33PAoHA9A2gcLn8H+N9weRTwx4Tj7wJGhsvHEIxFdES6P2+90vtqcENMSIPxubtnFa+YWRPgLjM7E/iK4JvwccCOhGPeBKaGZWe5+0oz+xbBZCWLw6E1mhJ8ky7PJDP7NcE4NVcSjF8z093/FcbwApADvAL83szuJmhOWlSN63oZuN/MmgGDgIXu/nnYHNXbzC4Oyx1NMFjcllLHtzCzleH1bwDmJpR/wsy6EAyz0KSC9/8ucKGZ3RKuNwfah+eSmFIikPriEqA10M/d91swomjzxALuvjBMFOcD08zsXmA3MNfdRyTxHj939+eLV8zs7PIKufs/LZjr4Dzgt2b2mrtPSOYi3L3IzBYA3wOGEUy0AsFsU9e5+6tVnOJzd88ys8MJxt+5FniAYAKe+e4+NOxYX1DB8QZc5O4bk4lX4kF9BFJfHA18FCaBgUCZOZctmIf5Q3d/BHiUYLq/14EzzKy4zf8IMzsxyfdcBHzfzA43syMImnUWmdl/Ap+5+9MEg/mVN2fs/rBmUp4ZBAOFFdcuIPijPrr4GDM7MXzPcnkw29z1wM3276HUi4ciHpVQdB9BE1mxV4HrLKweWTAqrcScEoHUF38Gss1sDXAZ8HY5Zc4CVpnZCoJv2/e7+06CP4zTzWw1QbNQt2Te0N3fIug7WEbQZ/Cou68ATgKWhU00twO/LefwKcDq4s7iUv4vwcRAf/Ng+kUIEtd64C0LJi1/mCpq7GEsqwkmZpkI/Hd47YnHzQd6FHcWE9QcmoSxrQvXJeZ0+6iISMypRiAiEnNKBCIiMadEICISc0oEIiIxp0QgIhJzSgQiIjGnRCAiEnP/H5nebBfRVD/JAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr, tpr, thresholds = metrics.roc_curve(y_test_spatial, y_pred)\n",
    "roc_auc = metrics.auc(fpr, tpr)\n",
    "display = metrics.RocCurveDisplay(fpr=fpr, tpr=tpr, roc_auc=roc_auc,\n",
    "                                   estimator_name='wv42 ROC curve')\n",
    "display.plot()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "71b37c8f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def draw_confusion_matrix(y, yhat, classes):\n",
    "    plt.cla()\n",
    "    plt.clf()\n",
    "    matrix = confusion_matrix(y, yhat)\n",
    "    plt.imshow(matrix, interpolation='nearest', cmap=plt.cm.Blues)\n",
    "    plt.title(\"wv42 confusion matrix\")\n",
    "    plt.colorbar()\n",
    "    num_classes = len(classes)\n",
    "    plt.xticks(np.arange(num_classes), classes, rotation=90)\n",
    "    plt.yticks(np.arange(num_classes), classes)\n",
    "\n",
    "    fmt = 'd'\n",
    "    thresh = matrix.max() / 2.\n",
    "    for i, j in itertools.product(range(matrix.shape[0]), range(matrix.shape[1])):\n",
    "        plt.text(j, i, format(matrix[i, j], fmt),\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if matrix[i, j] > thresh else \"black\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "91d2d582",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_confusion_matrix(y_test_spatial, y_pred_round, ['Glacial', 'Thermokarst'])\n",
    "fig = plt.ylabel('True label')\n",
    "fig = plt.xlabel('Predicted label')\n",
    "fig = plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "9fbb72c2",
   "metadata": {},
   "outputs": [],
   "source": [
    "#for i in range(0, 9):\n",
    "#    X_test_spatial_shuf = \n",
    "\n",
    "X_random = X_test_spatial\n",
    "\n",
    "X_test_spatial_approx = X_test_spatial[:, :, :, 0:9]\n",
    "\n",
    "X_test_spatial_approx = X_test_spatial_approx.reshape(1761264)\n",
    "\n",
    "np.random.shuffle(X_test_spatial_approx)\n",
    "\n",
    "X_test_spatial_shuf = X_test_spatial_approx.reshape(151, 36, 36, 9)\n",
    "\n",
    "X_random[:, :, :, 0:9] = X_test_spatial_shuf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "ef887306",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 1. 1. 1. 0. 1. 0. 0. 1. 0. 1. 0. 0. 0. 0. 1. 0. 1. 0. 1. 0. 0. 0. 1.\n",
      " 1. 1. 0. 1. 1. 0. 1. 0. 1. 0. 0. 1. 0. 1. 0. 0. 1. 1. 1. 1. 0. 0. 0. 1.\n",
      " 0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 1. 1. 1. 1. 0. 0. 1. 0.\n",
      " 0. 1. 0. 1. 0. 0. 0. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 0. 0. 1. 1. 1. 0. 1.\n",
      " 0. 0. 1. 1. 1. 0. 0. 1. 0. 0. 1. 0. 1. 0. 0. 0. 1. 0. 1. 0. 0. 1. 0. 0.\n",
      " 0. 0. 0. 0. 0. 1. 1. 0. 0. 1. 1. 0. 1. 0. 0. 1. 1. 1. 1. 0. 1. 1. 0. 1.\n",
      " 0. 1. 1. 1. 0. 1. 1.]\n"
     ]
    }
   ],
   "source": [
    "y_pred = model.predict([X_random, X_test_feat], )\n",
    "y_pred_round = np.rint(y_pred)\n",
    "\n",
    "y_pred_round = y_pred_round.reshape(y_pred_round.shape[0])\n",
    "\n",
    "print(y_pred_round)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "3618f6a5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "precision:  0.8732394366197183\n",
      "recall:  0.8266666666666667\n",
      "f1 score:  0.8493150684931506\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import precision_score, recall_score, f1_score\n",
    "\n",
    "print(\"precision: \", precision_score(y_test_spatial, y_pred_round))\n",
    "print(\"recall: \", recall_score(y_test_spatial, y_pred_round))\n",
    "print(\"f1 score: \", f1_score(y_test_spatial, y_pred_round))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "7b1edd3c",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr, tpr, thresholds = metrics.roc_curve(y_test_spatial, y_pred)\n",
    "roc_auc = metrics.auc(fpr, tpr)\n",
    "display = metrics.RocCurveDisplay(fpr=fpr, tpr=tpr, roc_auc=roc_auc,\n",
    "                                   estimator_name='wv42 with noise ROC curve')\n",
    "display.plot()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "dc234bbe",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def draw_confusion_matrix(y, yhat, classes):\n",
    "    plt.cla()\n",
    "    plt.clf()\n",
    "    matrix = confusion_matrix(y, yhat)\n",
    "    plt.imshow(matrix, interpolation='nearest', cmap=plt.cm.Blues)\n",
    "    plt.title(\"wv42 with noise confusion matrix\")\n",
    "    plt.colorbar()\n",
    "    num_classes = len(classes)\n",
    "    plt.xticks(np.arange(num_classes), classes, rotation=90)\n",
    "    plt.yticks(np.arange(num_classes), classes)\n",
    "\n",
    "    fmt = 'd'\n",
    "    thresh = matrix.max() / 2.\n",
    "    for i, j in itertools.product(range(matrix.shape[0]), range(matrix.shape[1])):\n",
    "        plt.text(j, i, format(matrix[i, j], fmt),\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if matrix[i, j] > thresh else \"black\")\n",
    "\n",
    "draw_confusion_matrix(y_test_spatial, y_pred_round, ['Glacial', 'Thermokarst'])\n",
    "fig = plt.ylabel('True label')\n",
    "fig = plt.xlabel('Predicted label')\n",
    "fig = plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "c09135db",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[-0.1273396   0.07012251 -0.12956372 ...  0.08088556 -0.06657421\n",
      "   0.14043139]\n",
      " [-0.08939543 -0.06321409 -0.03873656 ... -0.1427931  -0.03523848\n",
      "   0.06656288]\n",
      " [ 0.00219476  0.1111448   0.01054446 ...  0.1299481   0.07791828\n",
      "  -0.10399927]\n",
      " ...\n",
      " [-0.0381113   0.05041389 -0.00331391 ...  0.02168828  0.03949481\n",
      "  -0.09638578]\n",
      " [-0.0563642   0.0402343  -0.03405955 ... -0.02447303  0.01996617\n",
      "   0.09234674]\n",
      " [ 0.090397   -0.07663652  0.01945399 ... -0.13113566  0.07702585\n",
      "  -0.13278471]] 161\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[(3, 0.0795249193906784),\n",
       " (14, 0.07820068299770355),\n",
       " (13, 0.07737189531326294),\n",
       " (6, 0.07539163529872894),\n",
       " (8, 0.07528669387102127),\n",
       " (10, 0.07455205172300339),\n",
       " (9, 0.07236652076244354),\n",
       " (5, 0.07186135649681091),\n",
       " (15, 0.07070493698120117),\n",
       " (2, 0.07013995200395584),\n",
       " (11, 0.06977716833353043),\n",
       " (4, 0.06963646411895752),\n",
       " (0, 0.06919176131486893),\n",
       " (16, 0.0687849149107933),\n",
       " (1, 0.06864010542631149),\n",
       " (12, 0.06780591607093811),\n",
       " (7, 0.06549623608589172)]"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dense1_weights = model.layers[10].get_weights()[0]\n",
    "\n",
    "print(dense1_weights, len(dense1_weights))\n",
    "\n",
    "last_17 = abs(dense1_weights[-17:])\n",
    "\n",
    "feature_weight = []\n",
    "\n",
    "for i in range(len(last_17)):\n",
    "\n",
    "    hold_weights = last_17[i, :]\n",
    "    hold_sum = hold_weights.sum()\n",
    "    hold_avg = hold_sum / 128\n",
    "    feature_weight.append(hold_avg)\n",
    "    \n",
    "feature_weight_sort = sorted(enumerate(feature_weight), key = lambda x: x[1], reverse = True)\n",
    "\n",
    "feature_weight_sort"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "0a276d56",
   "metadata": {},
   "outputs": [],
   "source": [
    "names = \\\n",
    "['2 metre temperature',\n",
    " 'Evaporation',\n",
    " 'Evaporation from bare soil',\n",
    " 'Lake bottom temperature',\n",
    " 'Lake ice surface temperature',\n",
    " 'Lake ice total depth',\n",
    " 'Lake shape factor',\n",
    " 'Leaf area index, high vegetation',\n",
    " 'Leaf area index, low vegetation',\n",
    " 'Snow albedo',\n",
    " 'Snow cover',\n",
    " 'Snow depth',\n",
    " 'Snow evaporation',\n",
    " 'Snowmelt',\n",
    " 'Soil temperature level 1',\n",
    "#'Surface pressure',\n",
    " 'Surface runoff',\n",
    "#'Surface solar radiation downwards',\n",
    " 'Total precipitation']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "1e40ddd4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Lake bottom temperature\n",
      "Soil temperature level 1\n",
      "Snowmelt\n",
      "Lake shape factor\n",
      "Leaf area index, low vegetation\n",
      "Snow cover\n",
      "Snow albedo\n",
      "Lake ice total depth\n",
      "Surface runoff\n",
      "Evaporation from bare soil\n",
      "Snow depth\n",
      "Lake ice surface temperature\n",
      "2 metre temperature\n",
      "Total precipitation\n",
      "Evaporation\n",
      "Snow evaporation\n",
      "Leaf area index, high vegetation\n"
     ]
    }
   ],
   "source": [
    "feature_weight_sort = np.asarray(feature_weight_sort)\n",
    "list_best = feature_weight_sort[:, 0]\n",
    "list_best = list_best.astype('int')\n",
    "\n",
    "for i in list_best:\n",
    "    print(names[i])"
   ]
  }
 ],
 "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}