main_validation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2020 Global Flood Validation Metrics \n",
    "## For Submission\n",
    "Written in Python 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from pandas import DataFrame\n",
    "import numpy as np\n",
    "import os\n",
    "import glob\n",
    "import datetime\n",
    "import matplotlib as plt\n",
    "import datetime as datetime  \n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Step 1: Calculate erorr metrics for each flood"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run this section if you need to recalculate error metrics for each flood; otherwise skip to Step 2 to visualize and analyze"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Read and Examine Data Structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/ckuhn/Documents/Coauthor/GFD_Tellman/data\n"
     ]
    }
   ],
   "source": [
    "%cd /Users/ckuhn/Documents/Coauthor/GFD_Tellman/data/\n",
    "raw = pd.read_csv('gfd_validation_points_2018_12_17.csv',index_col=None, header=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "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>Unnamed: 0</th>\n",
       "      <th>Unnamed: 0.1</th>\n",
       "      <th>Unnamed: 0.1.1</th>\n",
       "      <th>B1</th>\n",
       "      <th>B2</th>\n",
       "      <th>B3</th>\n",
       "      <th>B4</th>\n",
       "      <th>B5</th>\n",
       "      <th>B7</th>\n",
       "      <th>MNDWI</th>\n",
       "      <th>...</th>\n",
       "      <th>pixel_lon</th>\n",
       "      <th>point_lat</th>\n",
       "      <th>point_lon</th>\n",
       "      <th>random</th>\n",
       "      <th>std_2day</th>\n",
       "      <th>std_3day</th>\n",
       "      <th>otsu_2day</th>\n",
       "      <th>otsu_3day</th>\n",
       "      <th>strata</th>\n",
       "      <th>validation</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>792.15651</td>\n",
       "      <td>896.579047</td>\n",
       "      <td>761.223987</td>\n",
       "      <td>1833.576595</td>\n",
       "      <td>980.842213</td>\n",
       "      <td>516.20095</td>\n",
       "      <td>-0.044882</td>\n",
       "      <td>...</td>\n",
       "      <td>87.277107</td>\n",
       "      <td>25.681778</td>\n",
       "      <td>87.277107</td>\n",
       "      <td>0.853473</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  Unnamed: 0.1  Unnamed: 0.1.1         B1          B2  \\\n",
       "0           0             0               0  792.15651  896.579047   \n",
       "\n",
       "           B3           B4          B5         B7     MNDWI     ...      \\\n",
       "0  761.223987  1833.576595  980.842213  516.20095 -0.044882     ...       \n",
       "\n",
       "   pixel_lon  point_lat  point_lon    random  std_2day  std_3day  otsu_2day  \\\n",
       "0  87.277107  25.681778  87.277107  0.853473       0.0       0.0        0.0   \n",
       "\n",
       "   otsu_3day  strata  validation  \n",
       "0        0.0       0         0.0  \n",
       "\n",
       "[1 rows x 27 columns]"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# dataframe first row\n",
    "raw.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(33141, 27)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Index(['Unnamed: 0', 'Unnamed: 0.1', 'Unnamed: 0.1.1', 'B1', 'B2', 'B3', 'B4',\n",
       "       'B5', 'B7', 'MNDWI', 'NDVI', 'analyst', 'date', 'dfoID', 'id', 'lon',\n",
       "       'pixel_lat', 'pixel_lon', 'point_lat', 'point_lon', 'random',\n",
       "       'std_2day', 'std_3day', 'otsu_2day', 'otsu_3day', 'strata',\n",
       "       'validation'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# dataframe shape\n",
    "print(raw.shape)\n",
    "\n",
    "# dataframe column headers\n",
    "raw.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "no. of unique floods: 123\n",
      "no of validation points: 33141\n",
      "representative breakdown:\n",
      "no of dry points: 8078\n",
      "no of permanent water points: 8107\n",
      "no of flooded points: 16952\n"
     ]
    }
   ],
   "source": [
    "# Validation** = student classification  \n",
    "# Strata** = DFO classification\n",
    "\n",
    "print('no. of unique floods:', len(raw['dfoID'].unique()))\n",
    "print('no of validation points:', len(raw))\n",
    "\n",
    "print('representative breakdown:')\n",
    "print('no of dry points:', len(raw.loc[raw['std_2day'] == 0]))\n",
    "print('no of permanent water points:', len(raw.loc[raw['std_2day'] == 1]))\n",
    "print('no of flooded points:', len(raw.loc[raw['std_2day'] == 2]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Drop floods with fewer than 100 points"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df = raw.groupby('dfoID').filter(lambda g: g.dfoID.count() >= 100)\n",
    "#df.dfoID.unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Drop NA values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Pixels got labeled \"NAN\" were coded as either -99 or 99 \n",
    "df = df.loc[df['validation'] > -1]\n",
    "df = df.loc[df['validation'] < 99]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculate error metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Background\n",
    "\n",
    "The points were stratified into three classes or strata. They are 50% not flooded (permanent water + dry) and 50% flooded. \n",
    "- not flooded/dry (0)\n",
    "- permanent water (1)\n",
    "- flooded (2)\n",
    "\n",
    "Four methods were used: \n",
    "- standard approach, 2 days\n",
    "- standard approach, 3 days\n",
    "- otsu, 2 days\n",
    "- otsu 3 days\n",
    "\n",
    "\n",
    "The **Validation** column is the student classification    \n",
    "The **Strata** column is the DFO classification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Select and prepare the method to analyze"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Isolate each of the methods\n",
    "st2 = df.filter(['dfoID', 'std_2day','strata', 'validation' ], axis=1)\n",
    "st3 = df.filter(['dfoID', 'std_3day','strata', 'validation' ], axis=1)\n",
    "ot2 = df.filter(['dfoID', 'otsu_2day','strata', 'validation' ], axis=1)\n",
    "ot3 = df.filter(['dfoID', 'otsu_3day','strata', 'validation' ], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "model_score\n",
      "0.0     7424\n",
      "1.0     7481\n",
      "2.0    15628\n",
      "Name: dfoID, dtype: int64\n",
      "student_score\n",
      "0.0    11497\n",
      "1.0    19036\n",
      "Name: dfoID, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# # Select which collection to analyze: \n",
    "\n",
    "# Create a label for the method to use in filenaming\n",
    "method = st2\n",
    "method_name = 'std_2day'\n",
    "\n",
    "method['model_score'] = method[method_name]\n",
    "method['student_score'] = method['validation']\n",
    "\n",
    "print(method.groupby('model_score')['dfoID'].count())\n",
    "# Remember: strata = 0 = dry, 1 = permanent water, 2 = flood\n",
    "print(method.groupby(['student_score'])['dfoID'].count())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculate Confusion Matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### matrix components\n",
    "\n",
    "\n",
    "- true positive = Validation (1) + Strata(4) = 5\n",
    "- true negative = Validation (0) + Strata (0) =  0 \n",
    "- false positive = Validation (0) + Strata (4) = 4\n",
    "- false negative = Validation (1) + Strata (0) = 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Calculate  metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "method['model_score'] = method['model_score'].multiply(2)\n",
    "method['model_score'].value_counts()\n",
    "# 0 - dry, 2 - permanent water - 4 - floods\n",
    "# 0 - dry, 1 - flooded\n",
    "method['student_score'].value_counts()\n",
    "# Add them together\n",
    "method['score']  = method['model_score'] + method['student_score'] \n",
    "method['score'].value_counts()\n",
    "\n",
    "# True Positive = Validation (1) + Strata (4)\n",
    "tp = len(method.loc[method['score']== 5])\n",
    "# True Negative = Validation (0) + Strata (0)\n",
    "tn = len(method.loc[method['score']== 0])\n",
    "# False Positive = Validation (0) + Strata (4)\n",
    "fp = len(method.loc[method['score']== 4])\n",
    "# False Negative = Validation (1) + Strata (0)\n",
    "fn = len(method.loc[method['score']== 1])\n",
    "\n",
    "\n",
    "precision = tp / (tp + fp)\n",
    "recall  = tp / (tp + fn)  # % of flooded pixels it ID'ed correctly\n",
    "accuracy = (tn + tp) / (tn+tp+fp+fn)\n",
    "commission = fp / (tp + fp)  # modeled positives over all true positives\n",
    "omission = fn / (tp + fn)\n",
    "\n",
    "total = tn+tp+fn+fp\n",
    "print('True Negatives(%)',tn/total) #32   Sep 2018 updates: 30%\n",
    "print('True Positives(%)',tp/total) #52   Sep 2018 updates: 52%\n",
    "print('False Positives(%)',fp/total) #15  Sep 2018 updates: 16%\n",
    "print('False Negatives(%)',fn/total) # <0.01  Sep 2018 updates: 1.6%\n",
    "print(total)\n",
    "print(accuracy)\n",
    "print(commission)\n",
    "print(omission)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "col_names =  ['Method', 'tp', 'tn','fp','fn',\n",
    "             'precision','recall','overall_accuracy',\n",
    "             'commission','omission'] ### Column names for making data frames for each flood\n",
    "\n",
    "my_df  = pd.DataFrame(columns = col_names)## Make empty data frame to put values for each sampling level in"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create new dataframe for that method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "std_2day = [method_name, tp, tn, fp, fn, precision, recall, accuracy, commission, omission]\n",
    "#std_3day = [method_name, tp, tn, fp, fn, precision, recall, accuracy, commission, omission]\n",
    "#otsu_2day = [method_name, tp, tn, fp, fn, precision, recall, accuracy, commission, omission]\n",
    "#otsu_3day = [method_name, tp, tn, fp, fn, precision, recall, accuracy, commission, omission]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "std2_short  = pd.DataFrame([std_2day], columns = col_names)## Make empty data frame to put values for each sampling level in\n",
    "# std3_short  = pd.DataFrame([std_3day], columns = col_names)## Make empty data frame to put values for each sampling level in\n",
    "# ot2_short  = pd.DataFrame([otsu_2day], columns = col_names)## Make empty data frame to put values for each sampling level in\n",
    "# ot3_short  = pd.DataFrame([otsu_3day], columns = col_names)## Make empty data frame to put values for each sampling level in"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## Optional step: add dataframes together\n",
    "# all_data = std2_short.append(std3_short).append(ot3_short).append(ot2_short)\n",
    "# all_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Step 2: Plot  error metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/ckuhn/Documents/Coauthor/GFD_Tellman/data\n"
     ]
    },
    {
     "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>Flood</th>\n",
       "      <th>Method</th>\n",
       "      <th>Latitude</th>\n",
       "      <th>Longitude</th>\n",
       "      <th>tp</th>\n",
       "      <th>tn</th>\n",
       "      <th>fp</th>\n",
       "      <th>fn</th>\n",
       "      <th>fp_jrc</th>\n",
       "      <th>tp_jrc</th>\n",
       "      <th>precision</th>\n",
       "      <th>recall</th>\n",
       "      <th>overall_accuracy</th>\n",
       "      <th>commission</th>\n",
       "      <th>omission</th>\n",
       "      <th>counts</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4098</td>\n",
       "      <td>std_3day</td>\n",
       "      <td>24.662609</td>\n",
       "      <td>87.105573</td>\n",
       "      <td>116.0</td>\n",
       "      <td>58.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>55.0</td>\n",
       "      <td>0.928</td>\n",
       "      <td>0.983051</td>\n",
       "      <td>0.940541</td>\n",
       "      <td>0.072</td>\n",
       "      <td>0.016949</td>\n",
       "      <td>245.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4098</td>\n",
       "      <td>otsu_3day</td>\n",
       "      <td>24.662609</td>\n",
       "      <td>87.105573</td>\n",
       "      <td>116.0</td>\n",
       "      <td>58.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>55.0</td>\n",
       "      <td>0.928</td>\n",
       "      <td>0.983051</td>\n",
       "      <td>0.940541</td>\n",
       "      <td>0.072</td>\n",
       "      <td>0.016949</td>\n",
       "      <td>245.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Flood     Method   Latitude  Longitude     tp    tn   fp   fn  fp_jrc  \\\n",
       "0   4098   std_3day  24.662609  87.105573  116.0  58.0  9.0  2.0     5.0   \n",
       "1   4098  otsu_3day  24.662609  87.105573  116.0  58.0  9.0  2.0     5.0   \n",
       "\n",
       "   tp_jrc  precision    recall  overall_accuracy  commission  omission  counts  \n",
       "0    55.0      0.928  0.983051          0.940541       0.072  0.016949   245.0  \n",
       "1    55.0      0.928  0.983051          0.940541       0.072  0.016949   245.0  "
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%cd '/Users/ckuhn/Documents/Coauthor/GFD_Tellman/data/'\n",
    "my_df = pd.read_csv('gfd_validation_metrics.csv').drop(['Unnamed: 0', 'MNDWI', 'NDVI'], axis=1)\n",
    "my_df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dfmelted = pd.melt(my_df, id_vars = ['Flood', 'Method', 'counts'], var_name = 'metric', value_name = 'value')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "error = dfmelted.loc[dfmelted['metric'].isin(['commission','omission'])]\n",
    "overall = dfmelted.loc[dfmelted['metric'].isin(['overall_accuracy','precision','recall' ])]\n",
    "basics = dfmelted.loc[dfmelted['metric'].isin(['tp','tn','fp', 'fn' ])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "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></th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Method</th>\n",
       "      <th>metric</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">otsu_2day</th>\n",
       "      <th>overall_accuracy</th>\n",
       "      <td>0.818962</td>\n",
       "      <td>0.135054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.773237</td>\n",
       "      <td>0.212571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.918766</td>\n",
       "      <td>0.121110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">otsu_3day</th>\n",
       "      <th>overall_accuracy</th>\n",
       "      <td>0.804849</td>\n",
       "      <td>0.123519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.788816</td>\n",
       "      <td>0.207127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.845200</td>\n",
       "      <td>0.170417</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">std_2day</th>\n",
       "      <th>overall_accuracy</th>\n",
       "      <td>0.823720</td>\n",
       "      <td>0.152845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.770245</td>\n",
       "      <td>0.215143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.965914</td>\n",
       "      <td>0.055223</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">std_3day</th>\n",
       "      <th>overall_accuracy</th>\n",
       "      <td>0.826149</td>\n",
       "      <td>0.145905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.772862</td>\n",
       "      <td>0.214098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.961809</td>\n",
       "      <td>0.059423</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                mean       std\n",
       "Method    metric                              \n",
       "otsu_2day overall_accuracy  0.818962  0.135054\n",
       "          precision         0.773237  0.212571\n",
       "          recall            0.918766  0.121110\n",
       "otsu_3day overall_accuracy  0.804849  0.123519\n",
       "          precision         0.788816  0.207127\n",
       "          recall            0.845200  0.170417\n",
       "std_2day  overall_accuracy  0.823720  0.152845\n",
       "          precision         0.770245  0.215143\n",
       "          recall            0.965914  0.055223\n",
       "std_3day  overall_accuracy  0.826149  0.145905\n",
       "          precision         0.772862  0.214098\n",
       "          recall            0.961809  0.059423"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "overall.groupby(['Method', 'metric'])['value'].agg(['mean', 'std'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "basics = basics.replace(['std_3day', 'otsu_2day', 'std_2day', 'otsu_3day'],['3Day Standard', '2Day Otsu', '2Day Standard', '3Day Otsu'])\n",
    "basics = basics.replace(['tp', 'tn', 'fp', 'fn'],['True Positive', 'True Negative', 'False Positive', 'False Negative'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/ckuhn/Documents/Coauthor/GFD_Tellman/figures\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def set_style():\n",
    "    plt.style.use(['seaborn-white', 'seaborn-talk'])\n",
    "    plt.rc(\"font\", family=\"Times New Roman\", size = 30)\n",
    "    plt.rc('xtick', labelsize=30) \n",
    "    plt.rc('ytick', labelsize=30) \n",
    "set_style()  \n",
    "\n",
    "order = 'True Positive', 'True Negative', 'False Positive', 'False Negative'\n",
    "\n",
    "g = sns.factorplot(kind='box',        # Boxplot\n",
    "               y='value',       # Y-axis - values for boxplot\n",
    "               x='metric',        # X-axis - first factor\n",
    "               hue = 'Method',\n",
    "               hue_order  = ['2Day Otsu', '2Day Standard', '3Day Otsu', '3Day Standard'],\n",
    "               data=basics,        # Dataframe \n",
    "               order = order,\n",
    "               size=8,            # Figure size (x100px)      \n",
    "               aspect=1.5, \n",
    "               palette= ['#ca0020', '#f4a582', '#0571b0', '#92c5de'],  # Width = size * aspect \n",
    "               legend_out=False)  # Make legend inside the plot\n",
    "#g.fig.get_axes()[0].set_yscale('log')\n",
    "\n",
    "#g.fig.text(0.75, 0.75,'n = 121 floods', fontsize=30, color = '#ff3d00') #add text\n",
    "\n",
    "g.set_xticklabels(rotation=30)\n",
    "\n",
    "plt.ylabel('Count', fontsize = 45)\n",
    "plt.xlabel('', fontsize = 45)\n",
    "plt.title('', fontsize = 45)\n",
    "#plt.axhline(y=0, color='grey', ls='--')\n",
    "#plt.legend(loc='upper left', fontsize=45)\n",
    "plt.legend()\n",
    "plt.tight_layout()\n",
    "\n",
    "%cd /Users/ckuhn/Documents/Coauthor/GFD_Tellman/figures/\n",
    "plt.savefig('GFD_basics.png', dpi = 300)\n",
    "\n",
    "# plt.show()\n",
    "\n",
    "#http://queirozf.com/entries/matplotlib-pyplot-by-example#change-tick-label-rotation # ADD LABELS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Method         metric        \n",
       "2Day Otsu      False Negative      6.974359\n",
       "               False Positive     27.709402\n",
       "               True Negative      61.239316\n",
       "               True Positive      98.085470\n",
       "2Day Standard  False Negative      3.228814\n",
       "               False Positive     30.381356\n",
       "               True Negative      58.305085\n",
       "               True Positive     101.296610\n",
       "3Day Otsu      False Negative     14.025641\n",
       "               False Positive     23.418803\n",
       "               True Negative      65.529915\n",
       "               True Positive      91.034188\n",
       "3Day Standard  False Negative      3.661017\n",
       "               False Positive     29.542373\n",
       "               True Negative      59.144068\n",
       "               True Positive     100.864407\n",
       "Name: value, dtype: float64"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "basics.groupby(['Method', 'metric'])['value'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "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></th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Method</th>\n",
       "      <th>metric</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">otsu_2day</th>\n",
       "      <th>commission</th>\n",
       "      <td>0.226763</td>\n",
       "      <td>0.212571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>omission</th>\n",
       "      <td>0.081234</td>\n",
       "      <td>0.121110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">otsu_3day</th>\n",
       "      <th>commission</th>\n",
       "      <td>0.211184</td>\n",
       "      <td>0.207127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>omission</th>\n",
       "      <td>0.154800</td>\n",
       "      <td>0.170417</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">std_2day</th>\n",
       "      <th>commission</th>\n",
       "      <td>0.229755</td>\n",
       "      <td>0.215143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>omission</th>\n",
       "      <td>0.034086</td>\n",
       "      <td>0.055223</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">std_3day</th>\n",
       "      <th>commission</th>\n",
       "      <td>0.227138</td>\n",
       "      <td>0.214098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>omission</th>\n",
       "      <td>0.038191</td>\n",
       "      <td>0.059423</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          mean       std\n",
       "Method    metric                        \n",
       "otsu_2day commission  0.226763  0.212571\n",
       "          omission    0.081234  0.121110\n",
       "otsu_3day commission  0.211184  0.207127\n",
       "          omission    0.154800  0.170417\n",
       "std_2day  commission  0.229755  0.215143\n",
       "          omission    0.034086  0.055223\n",
       "std_3day  commission  0.227138  0.214098\n",
       "          omission    0.038191  0.059423"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "error.groupby(['Method', 'metric'])['value'].agg(['mean', 'std'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "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></th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Method</th>\n",
       "      <th>metric</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">otsu_2day</th>\n",
       "      <th>overall_accuracy</th>\n",
       "      <td>0.818962</td>\n",
       "      <td>0.135054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.773237</td>\n",
       "      <td>0.212571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.918766</td>\n",
       "      <td>0.121110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">otsu_3day</th>\n",
       "      <th>overall_accuracy</th>\n",
       "      <td>0.804849</td>\n",
       "      <td>0.123519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.788816</td>\n",
       "      <td>0.207127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.845200</td>\n",
       "      <td>0.170417</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">std_2day</th>\n",
       "      <th>overall_accuracy</th>\n",
       "      <td>0.823720</td>\n",
       "      <td>0.152845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.770245</td>\n",
       "      <td>0.215143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.965914</td>\n",
       "      <td>0.055223</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">std_3day</th>\n",
       "      <th>overall_accuracy</th>\n",
       "      <td>0.826149</td>\n",
       "      <td>0.145905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.772862</td>\n",
       "      <td>0.214098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.961809</td>\n",
       "      <td>0.059423</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                mean       std\n",
       "Method    metric                              \n",
       "otsu_2day overall_accuracy  0.818962  0.135054\n",
       "          precision         0.773237  0.212571\n",
       "          recall            0.918766  0.121110\n",
       "otsu_3day overall_accuracy  0.804849  0.123519\n",
       "          precision         0.788816  0.207127\n",
       "          recall            0.845200  0.170417\n",
       "std_2day  overall_accuracy  0.823720  0.152845\n",
       "          precision         0.770245  0.215143\n",
       "          recall            0.965914  0.055223\n",
       "std_3day  overall_accuracy  0.826149  0.145905\n",
       "          precision         0.772862  0.214098\n",
       "          recall            0.961809  0.059423"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "overall.groupby(['Method', 'metric'])['value'].agg(['mean', 'std'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Step 3: Test Number of Points"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This step takes a subsample of floods we used for testing and determines how the accuracy changes as a function of the number of points tha tyou sample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/ckuhn/Documents/Coauthor/GFD_Tellman\n"
     ]
    }
   ],
   "source": [
    "%cd '/Users/ckuhn/Documents/Coauthor/GFD_Tellman/'\n",
    "df = pd.read_csv('gfd_validation_sensitivity.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "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>Unnamed: 0</th>\n",
       "      <th>B1</th>\n",
       "      <th>B2</th>\n",
       "      <th>B3</th>\n",
       "      <th>B4</th>\n",
       "      <th>B5</th>\n",
       "      <th>B7</th>\n",
       "      <th>MNDWI</th>\n",
       "      <th>NDVI</th>\n",
       "      <th>analyst</th>\n",
       "      <th>date</th>\n",
       "      <th>dfoID</th>\n",
       "      <th>id</th>\n",
       "      <th>lat</th>\n",
       "      <th>lon</th>\n",
       "      <th>random</th>\n",
       "      <th>strata</th>\n",
       "      <th>validation</th>\n",
       "      <th>score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>675.522719</td>\n",
       "      <td>1128.643606</td>\n",
       "      <td>2015.265677</td>\n",
       "      <td>2436.991734</td>\n",
       "      <td>3702.040962</td>\n",
       "      <td>2753.348602</td>\n",
       "      <td>-0.365915</td>\n",
       "      <td>0.095263</td>\n",
       "      <td>HT</td>\n",
       "      <td>5/10/18</td>\n",
       "      <td>2143</td>\n",
       "      <td>0</td>\n",
       "      <td>-21.762688</td>\n",
       "      <td>123.87866</td>\n",
       "      <td>0.341315</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0          B1           B2           B3           B4           B5  \\\n",
       "0           0  675.522719  1128.643606  2015.265677  2436.991734  3702.040962   \n",
       "\n",
       "            B7     MNDWI      NDVI analyst     date  dfoID  id        lat  \\\n",
       "0  2753.348602 -0.365915  0.095263      HT  5/10/18   2143   0 -21.762688   \n",
       "\n",
       "         lon    random  strata  validation  score  \n",
       "0  123.87866  0.341315       0           0      0  "
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw = df.copy()\n",
    "raw.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# These are the ones we are going to drop \n",
    "raw3 = raw.groupby('dfoID').filter(lambda g: g.dfoID.count() < 200)\n",
    "raw3.dfoID.unique()\n",
    "\n",
    "#From the first round of classifications\n",
    "raw = raw[raw.dfoID != 4507]\n",
    "raw = raw[raw.dfoID != 2586]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "floods = list(set(raw.dfoID)) ## Get a list of all the unique values of floods\n",
    "sampling_levels = np.arange(0, 400, 1)\n",
    "\n",
    "#sampling_levels = [10,20,30,40,50,60,100] ### Set levels you want to sample at \n",
    "\n",
    "col_names =  ['Flood','NumofPoints', 'tp', 'tn','fp','fn',\n",
    "             'precision','recall','overall_accuracy',\n",
    "             'commission','omission'] ### Column names for making data frames for each flood\n",
    "\n",
    "my_df  = pd.DataFrame(columns = col_names)## Make empty data frame to put values for each sampling level in\n",
    "\n",
    "for i in floods:  ### Loop through floods\n",
    "    df_full = raw.loc[raw['dfoID']== i] ### subset data frame by single flood\n",
    "    for j in sampling_levels: ## Loop through sampling levels\n",
    "        df = df_full.sample(j) ##Apparently pandas has a built-in dataframe random sampler!\n",
    "        tp = float(len(df.loc[df['score']== 5])) \n",
    "        tn = float(len(df.loc[df['score']== 0]))\n",
    "        fp = float(len(df.loc[df['score']== 4]))\n",
    "        fn = float(len(df.loc[df['score']== 1]))\n",
    "        ### Calculating values!\n",
    "        if tp == 0: ###Hmm, sometimes there are no tp scores, so you get a zero divide error\n",
    "            continue ###I've put this in for now but it's gonna bite your butt eventually\n",
    "        precision = tp / (tp + fp)\n",
    "        recall  = tp / (tp + fn)  # % of flooded pixels it ID'ed correctly\n",
    "        accuracy = (tn + tp) / (tn+tp+fp+fn)\n",
    "        commission = fp / (tp + fp)  # modeled positives over all true positives\n",
    "        omission = fn / (tp + fn)\n",
    "        ##add a new row to the dataframe based on this sampling level!\n",
    "        my_df.loc[len(my_df)] = [i,j, tp, tn, fp, fn, precision, recall, accuracy, commission, omission]\n",
    "\n",
    "        ###Now you should have a full dataframe with all your sampling levels for each flood\n",
    "        ###SO\n",
    "#my_df.to_csv('myprecious.csv', encoding='utf-8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "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>Flood</th>\n",
       "      <th>NumofPoints</th>\n",
       "      <th>tp</th>\n",
       "      <th>tn</th>\n",
       "      <th>fp</th>\n",
       "      <th>fn</th>\n",
       "      <th>precision</th>\n",
       "      <th>recall</th>\n",
       "      <th>overall_accuracy</th>\n",
       "      <th>commission</th>\n",
       "      <th>omission</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4163.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Flood  NumofPoints   tp   tn   fp   fn  precision  recall  \\\n",
       "0  4163.0          3.0  1.0  0.0  0.0  0.0        1.0     1.0   \n",
       "\n",
       "   overall_accuracy  commission  omission  \n",
       "0               1.0         0.0       0.0  "
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_df.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.17346292877842398\n",
      "0.025846575186897005\n",
      "0.8749003938035954\n"
     ]
    }
   ],
   "source": [
    "print(my_df.commission.mean())\n",
    "print(my_df.omission.mean())\n",
    "print(my_df.overall_accuracy.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def set_style():\n",
    "    plt.style.use(['seaborn-white', 'seaborn-talk'])\n",
    "    plt.rc(\"font\", family=\"Times New Roman\", size = 30)\n",
    "    plt.rc('xtick', labelsize=30) \n",
    "    plt.rc('ytick', labelsize=30) \n",
    "    plt.rc({'axes.titlesize': 'large'})\n",
    "set_style()  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0, 400)"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 748.8x514.8 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "my_df.groupby(['NumofPoints'])['overall_accuracy'].mean().plot(label='Overall Accuracy')\n",
    "\n",
    "plt.xlabel('Number of Points', fontsize=30)\n",
    "\n",
    "my_df.groupby(['NumofPoints'])['precision'].mean().plot(label='Precision')\n",
    "plt.xlabel('Number of Points', fontsize=30)\n",
    "\n",
    "my_df.groupby(['NumofPoints'])['recall'].mean().plot(label='Recall')\n",
    "plt.xlabel('Number of Points', fontsize=30)\n",
    "plt.ylabel ('Score', fontsize=40)\n",
    "plt.legend(title=\"Metric\")\n",
    "\n",
    "plt.axvline(250, color='grey', linestyle='--')\n",
    "\n",
    "plt.ylim(0.6,1.01)\n",
    "plt.xlim(0,400)\n",
    "\n",
    "# %cd /Users/ckuhn/Documents/Coauthor/DFO_Accuracy/figures/\n",
    "# plt.savefig('point_test_accuracy_n10_2019_revisions.png', dpi = 300)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Std. Dev of Accuracy @ any given number of points"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.09778886675976595"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = my_df.groupby(['NumofPoints'])['overall_accuracy'].agg(['mean','var', 'std']).reset_index()\n",
    "\n",
    "df.loc[df['NumofPoints']==250]['std'].mean()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}