models/knn_5country.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "toc": true
   },
   "source": [
    "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n",
    "<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Imports-and-Constants\" data-toc-modified-id=\"Imports-and-Constants-1\"><span class=\"toc-item-num\">1&nbsp;&nbsp;</span>Imports and Constants</a></span></li><li><span><a href=\"#Helper-Functions\" data-toc-modified-id=\"Helper-Functions-2\"><span class=\"toc-item-num\">2&nbsp;&nbsp;</span>Helper Functions</a></span></li><li><span><a href=\"#Load-DHS-Data\" data-toc-modified-id=\"Load-DHS-Data-3\"><span class=\"toc-item-num\">3&nbsp;&nbsp;</span>Load DHS Data</a></span></li><li><span><a href=\"#Load-LSMS-Data\" data-toc-modified-id=\"Load-LSMS-Data-4\"><span class=\"toc-item-num\">4&nbsp;&nbsp;</span>Load LSMS Data</a></span><ul class=\"toc-item\"><li><span><a href=\"#Neal-OOC-setup\" data-toc-modified-id=\"Neal-OOC-setup-4.1\"><span class=\"toc-item-num\">4.1&nbsp;&nbsp;</span>Neal OOC setup</a></span></li></ul></li><li><span><a href=\"#Neal-NL-and-Labels-vs.-new-NL-and-Labels\" data-toc-modified-id=\"Neal-NL-and-Labels-vs.-new-NL-and-Labels-5\"><span class=\"toc-item-num\">5&nbsp;&nbsp;</span>Neal NL and Labels vs. new NL and Labels</a></span></li><li><span><a href=\"#KNN\" data-toc-modified-id=\"KNN-6\"><span class=\"toc-item-num\">6&nbsp;&nbsp;</span>KNN</a></span><ul class=\"toc-item\"><li><span><a href=\"#Neal-OOC-setup\" data-toc-modified-id=\"Neal-OOC-setup-6.1\"><span class=\"toc-item-num\">6.1&nbsp;&nbsp;</span>Neal OOC setup</a></span></li><li><span><a href=\"#Label-vs.-NL\" data-toc-modified-id=\"Label-vs.-NL-6.2\"><span class=\"toc-item-num\">6.2&nbsp;&nbsp;</span>Label vs. NL</a></span></li><li><span><a href=\"#Neal-Label-vs.-Neal-NL\" data-toc-modified-id=\"Neal-Label-vs.-Neal-NL-6.3\"><span class=\"toc-item-num\">6.3&nbsp;&nbsp;</span>Neal Label vs. Neal NL</a></span></li><li><span><a href=\"#Neal-Label-vs.-NL\" data-toc-modified-id=\"Neal-Label-vs.-NL-6.4\"><span class=\"toc-item-num\">6.4&nbsp;&nbsp;</span>Neal Label vs. NL</a></span></li></ul></li></ul></div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Imports and Constants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import defaultdict\n",
    "import os\n",
    "import pickle\n",
    "import sys\n",
    "\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import sklearn.neighbors\n",
    "\n",
    "from batchers import batcher, dataset_constants\n",
    "from models.histograms import (\n",
    "    get_per_image_histograms,\n",
    "    plot_band_hists,\n",
    "    plot_label_hist,\n",
    "    split_nl_hist)\n",
    "from models.linear_model import ridge_cv\n",
    "from utils.analysis import calc_score, evaluate\n",
    "from utils.general import load_npz\n",
    "from utils.plot import scatter_preds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "FOLDS = ['A', 'B', 'C', 'D']\n",
    "SPLITS = ['train', 'val', 'test']\n",
    "\n",
    "MEANS = dataset_constants.MEANS_DICT['2009-17']\n",
    "STD_DEVS = dataset_constants.STD_DEVS_DICT['2009-17']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Helper Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def background_gradient(df, cmap, m=None, M=None):\n",
    "    '''\n",
    "    Args\n",
    "    - df: pd.DataFrame\n",
    "    - cmap: str, name of matplotlib colormap\n",
    "    - m: float, value corresponding to lowest color in cmap\n",
    "    - M: float, value corresponding to highest color in cmap\n",
    "\n",
    "    Returns\n",
    "    - styles_df: pd.DataFrame\n",
    "    '''\n",
    "    if m is None:\n",
    "        m = np.nanmin(df.values)\n",
    "    if M is None:\n",
    "        M = np.nanmax(df.values)\n",
    "    normed = np.clip((df.values - m) / (M - m), 0, 1)\n",
    "    normed[np.isnan(normed)] = 0\n",
    "    cmap = plt.cm.get_cmap(cmap)\n",
    "    styles = []\n",
    "    for rgba in cmap(normed.ravel()):\n",
    "        c = matplotlib.colors.rgb2hex(rgba)\n",
    "        if np.mean(rgba) < 0.75:\n",
    "            styles.append(f'background-color: {c}; color: white;')\n",
    "        else:\n",
    "            styles.append(f'background-color: {c};')\n",
    "    styles = np.asarray(styles).reshape(df.shape)\n",
    "    styles_df = pd.DataFrame(styles, index=df.index, columns=df.columns)\n",
    "    styles_df[df.isna()] = f'background-color: white'\n",
    "    return styles_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def neal_ooc(df, country_order):\n",
    "    '''\n",
    "    Args\n",
    "    - df: pd.DataFrame\n",
    "    - countries: list of countries\n",
    "    '''\n",
    "    knn = sklearn.neighbors.KNeighborsRegressor(100)\n",
    "    neal_ooc_df = pd.DataFrame(columns=['train_on', 'eval_on', 'r2'], dtype=np.float32)\n",
    "    neal_ooc_df.set_index(['train_on', 'eval_on'], inplace=True)\n",
    "    for train_c in country_order:\n",
    "        for test_c in country_order:\n",
    "            if test_c == train_c: continue\n",
    "\n",
    "            train_mask = (df['country'] == train_c)\n",
    "            test_mask = (df['country'] == test_c)\n",
    "            knn.fit(X=df.loc[train_mask, ['neal_nl']].values, y=df.loc[train_mask, 'neal_label'].values)\n",
    "            preds = knn.predict(X=df.loc[test_mask, ['neal_nl']].values)\n",
    "            neal_ooc_df.loc[(train_c, test_c), :] = calc_score(\n",
    "                df.loc[test_mask, 'neal_label'].values, preds, metric='r2')\n",
    "\n",
    "    rows = country_order\n",
    "    cols = rows[::-1]\n",
    "\n",
    "    return neal_ooc_df.stack().reset_index(level=2, drop=True).unstack('train_on').loc[rows, cols]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load DHS Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "NEAL_DIR = '/atlas/u/nj/dhs/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "surveys = [\n",
    "    'malawi_dhs_2010',\n",
    "    'nigeria_dhs_2013',\n",
    "    'rwanda_dhs_2010',\n",
    "    'tanzania_dhs_2010',\n",
    "    'uganda_dhs_2011',\n",
    "]\n",
    "\n",
    "COUNTRIES = ['malawi', 'nigeria', 'rwanda', 'tanzania', 'uganda']\n",
    "DMSP_COUNTRIES = ['malawi', 'rwanda', 'tanzania', 'uganda']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.DataFrame(columns=['lat', 'lon', 'country', 'year', 'neal_nl', 'neal_label', 'nl', 'label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "image_hists: dtype=int64, shape=(19669, 8, 102)\n",
      "labels: dtype=float32, shape=(19669,)\n",
      "locs: dtype=float32, shape=(19669, 2)\n",
      "years: dtype=int32, shape=(19669,)\n",
      "nls_center: dtype=float32, shape=(19669,)\n",
      "nls_mean: dtype=float32, shape=(19669,)\n"
     ]
    }
   ],
   "source": [
    "file_path = 'data/dhs_image_hists.npz'\n",
    "result = load_npz(file_path)\n",
    "image_hists = result['image_hists']\n",
    "labels = result['labels']\n",
    "locs = result['locs']\n",
    "years = result['years']\n",
    "nls_center = result['nls_center']\n",
    "nls_mean = result['nls_mean']\n",
    "\n",
    "dmsp_mask = years < 2012\n",
    "viirs_mask = ~dmsp_mask"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "loc_to_index = {\n",
    "    (lat, lon): i\n",
    "    for i, (lat, lon) in enumerate(locs)\n",
    "}\n",
    "assert len(loc_to_index) == len(locs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "for survey in surveys:\n",
    "    country, year = survey.split('_dhs_')\n",
    "    year = int(year)\n",
    "\n",
    "    folder = os.path.join(NEAL_DIR, survey, 'data')\n",
    "    image_counts = np.load(os.path.join(folder, 'cluster_image_counts.npy'))\n",
    "    mask = (image_counts >= 10)\n",
    "\n",
    "    clocs = np.load(os.path.join(folder, 'cluster_locs.npy'))\n",
    "    clocs = clocs[mask]\n",
    "\n",
    "    neal_nls = np.load(os.path.join(folder, 'cluster_nightlights.npy'))\n",
    "    neal_nls = neal_nls[mask]\n",
    "\n",
    "    neal_labels = np.load(os.path.join(folder, 'cluster_assets.npy'))\n",
    "    neal_labels = neal_labels[mask]\n",
    "\n",
    "    indices = [\n",
    "        loc_to_index[(lat, lon)]\n",
    "        for lat, lon in clocs.astype(np.float32)\n",
    "    ]\n",
    "    \n",
    "    cdf = pd.DataFrame(data={\n",
    "        'lat': clocs[:, 0],\n",
    "        'lon': clocs[:, 1],\n",
    "        'country': country,\n",
    "        'year': year,\n",
    "        'neal_nl': neal_nls,\n",
    "        'neal_label': neal_labels,\n",
    "        'nl': nls_mean[indices],\n",
    "        'label': labels[indices]\n",
    "    })\n",
    "    df = pd.concat([df, cdf])\n",
    "df.reset_index(inplace=True, drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>lat</th>\n",
       "      <th>lon</th>\n",
       "      <th>country</th>\n",
       "      <th>year</th>\n",
       "      <th>neal_nl</th>\n",
       "      <th>neal_label</th>\n",
       "      <th>nl</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-14.293995</td>\n",
       "      <td>34.468751</td>\n",
       "      <td>malawi</td>\n",
       "      <td>2010</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.170203</td>\n",
       "      <td>-0.173862</td>\n",
       "      <td>-0.476676</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-14.791538</td>\n",
       "      <td>35.033085</td>\n",
       "      <td>malawi</td>\n",
       "      <td>2010</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.137892</td>\n",
       "      <td>-0.173862</td>\n",
       "      <td>-0.556283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-13.435848</td>\n",
       "      <td>32.945216</td>\n",
       "      <td>malawi</td>\n",
       "      <td>2010</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.265330</td>\n",
       "      <td>-0.173862</td>\n",
       "      <td>-0.559472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3031</th>\n",
       "      <td>-0.796772</td>\n",
       "      <td>29.931162</td>\n",
       "      <td>uganda</td>\n",
       "      <td>2011</td>\n",
       "      <td>5.2</td>\n",
       "      <td>1.256519</td>\n",
       "      <td>-0.103255</td>\n",
       "      <td>0.685674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3032</th>\n",
       "      <td>-0.967814</td>\n",
       "      <td>29.988417</td>\n",
       "      <td>uganda</td>\n",
       "      <td>2011</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.419588</td>\n",
       "      <td>-0.173862</td>\n",
       "      <td>-0.651027</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3033</th>\n",
       "      <td>-0.688925</td>\n",
       "      <td>29.938125</td>\n",
       "      <td>uganda</td>\n",
       "      <td>2011</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.209665</td>\n",
       "      <td>-0.173862</td>\n",
       "      <td>-0.565248</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3034 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            lat        lon country  year  neal_nl  neal_label        nl  \\\n",
       "0    -14.293995  34.468751  malawi  2010      0.0    0.170203 -0.173862   \n",
       "1    -14.791538  35.033085  malawi  2010      0.0   -0.137892 -0.173862   \n",
       "2    -13.435848  32.945216  malawi  2010      0.0   -0.265330 -0.173862   \n",
       "...         ...        ...     ...   ...      ...         ...       ...   \n",
       "3031  -0.796772  29.931162  uganda  2011      5.2    1.256519 -0.103255   \n",
       "3032  -0.967814  29.988417  uganda  2011      0.0   -0.419588 -0.173862   \n",
       "3033  -0.688925  29.938125  uganda  2011      0.0   -0.209665 -0.173862   \n",
       "\n",
       "         label  \n",
       "0    -0.476676  \n",
       "1    -0.556283  \n",
       "2    -0.559472  \n",
       "...        ...  \n",
       "3031  0.685674  \n",
       "3032 -0.651027  \n",
       "3033 -0.565248  \n",
       "\n",
       "[3034 rows x 8 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "with pd.option_context('display.max_rows', 6):\n",
    "    display(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load LSMS Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "NEAL_LSMS_DIR = '/atlas/u/nj/lsms'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "lsms_surveys = [\n",
    "    'malawi_lsms_2013',\n",
    "    'nigeria_lsms_2012_2013',\n",
    "    'tanzania_lsms_2012_2013',\n",
    "    'uganda_lsms_2011_2012',\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>lat</th>\n",
       "      <th>lon</th>\n",
       "      <th>country</th>\n",
       "      <th>year</th>\n",
       "      <th>neal_nl</th>\n",
       "      <th>neal_label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-9.504460</td>\n",
       "      <td>33.230200</td>\n",
       "      <td>malawi</td>\n",
       "      <td>2013</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.460405</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-9.667920</td>\n",
       "      <td>33.393700</td>\n",
       "      <td>malawi</td>\n",
       "      <td>2013</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.946366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-9.809980</td>\n",
       "      <td>33.387300</td>\n",
       "      <td>malawi</td>\n",
       "      <td>2013</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.517353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1408</th>\n",
       "      <td>-0.925602</td>\n",
       "      <td>30.794159</td>\n",
       "      <td>uganda</td>\n",
       "      <td>2011</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.502750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1409</th>\n",
       "      <td>-0.757246</td>\n",
       "      <td>30.807199</td>\n",
       "      <td>uganda</td>\n",
       "      <td>2011</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.869549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1410</th>\n",
       "      <td>-0.151138</td>\n",
       "      <td>30.948854</td>\n",
       "      <td>uganda</td>\n",
       "      <td>2011</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.363786</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1411 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           lat        lon country  year  neal_nl  neal_label\n",
       "0    -9.504460  33.230200  malawi  2013      0.0    1.460405\n",
       "1    -9.667920  33.393700  malawi  2013      0.0    1.946366\n",
       "2    -9.809980  33.387300  malawi  2013      0.0    1.517353\n",
       "...        ...        ...     ...   ...      ...         ...\n",
       "1408 -0.925602  30.794159  uganda  2011      0.0    2.502750\n",
       "1409 -0.757246  30.807199  uganda  2011      0.0    5.869549\n",
       "1410 -0.151138  30.948854  uganda  2011      0.0    2.363786\n",
       "\n",
       "[1411 rows x 6 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def load_neal_lsms_data(lsms_surveys):\n",
    "    df = pd.DataFrame(columns=['lat', 'lon', 'country', 'year', 'neal_nl', 'neal_label'])\n",
    "    for survey in lsms_surveys:\n",
    "        country, years = survey.split('_lsms_')\n",
    "        year = int(years.split('_')[0])\n",
    "\n",
    "        folder = os.path.join(NEAL_LSMS_DIR, survey, 'data')\n",
    "        image_counts = np.load(os.path.join(folder, 'cluster_image_counts.npy'))\n",
    "        mask = (image_counts >= 10)\n",
    "\n",
    "        households = np.load(os.path.join(folder, 'cluster_households.npy'))\n",
    "        mask &= (households >= 2)\n",
    "\n",
    "        clocs = np.load(os.path.join(folder, 'cluster_locs.npy'))\n",
    "        clocs = clocs[mask]\n",
    "\n",
    "        neal_nls = np.load(os.path.join(folder, 'cluster_nightlights.npy'))\n",
    "        neal_nls = neal_nls[mask]\n",
    "\n",
    "        neal_labels = np.load(os.path.join(folder, 'cluster_consumptions.npy'))\n",
    "        neal_labels = neal_labels[mask]\n",
    "\n",
    "        cdf = pd.DataFrame(data={\n",
    "            'lat': clocs[:, 0],\n",
    "            'lon': clocs[:, 1],\n",
    "            'country': country,\n",
    "            'year': year,\n",
    "            'neal_nl': neal_nls,\n",
    "            'neal_label': neal_labels,\n",
    "        })\n",
    "        df = pd.concat([df, cdf])\n",
    "    df.reset_index(inplace=True, drop=True)\n",
    "    return df\n",
    "\n",
    "lsms_df = load_neal_lsms_data(lsms_surveys)\n",
    "with pd.option_context('display.max_rows', 6):\n",
    "    display(lsms_df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Neal OOC setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_622c9060_9861_11e9_83d4_39d9ea9de529row0_col0 {\n",
       "            background-color:  #92b4fe;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row0_col1 {\n",
       "            background-color:  #a5c3fe;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row0_col2 {\n",
       "            background-color:  #9bbcff;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row0_col3 {\n",
       "            background-color:  white;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row1_col0 {\n",
       "            background-color:  #f3c7b1;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row1_col1 {\n",
       "            background-color:  #ead5c9;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row1_col2 {\n",
       "            background-color:  white;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row1_col3 {\n",
       "            background-color:  #dedcdb;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row2_col0 {\n",
       "            background-color:  #c0282f;\n",
       "             color:  white;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row2_col1 {\n",
       "            background-color:  white;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row2_col2 {\n",
       "            background-color:  #cb3e38;\n",
       "             color:  white;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row2_col3 {\n",
       "            background-color:  #cf453c;\n",
       "             color:  white;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row3_col0 {\n",
       "            background-color:  white;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row3_col1 {\n",
       "            background-color:  #a5c3fe;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row3_col2 {\n",
       "            background-color:  #a6c4fe;\n",
       "        }    #T_622c9060_9861_11e9_83d4_39d9ea9de529row3_col3 {\n",
       "            background-color:  #8fb1fe;\n",
       "        }</style><table id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529\" ><thead>    <tr>        <th class=\"index_name level0\" >train_on</th>        <th class=\"col_heading level0 col0\" >nigeria</th>        <th class=\"col_heading level0 col1\" >tanzania</th>        <th class=\"col_heading level0 col2\" >uganda</th>        <th class=\"col_heading level0 col3\" >malawi</th>    </tr>    <tr>        <th class=\"index_name level0\" >eval_on</th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529level0_row0\" class=\"row_heading level0 row0\" >malawi</th>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row0_col0\" class=\"data row0 col0\" >0.28</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row0_col1\" class=\"data row0 col1\" >0.30</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row0_col2\" class=\"data row0 col2\" >0.29</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row0_col3\" class=\"data row0 col3\" >nan</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529level0_row1\" class=\"row_heading level0 row1\" >uganda</th>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row1_col0\" class=\"data row1 col0\" >0.42</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row1_col1\" class=\"data row1 col1\" >0.40</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row1_col2\" class=\"data row1 col2\" >nan</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row1_col3\" class=\"data row1 col3\" >0.38</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529level0_row2\" class=\"row_heading level0 row2\" >tanzania</th>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row2_col0\" class=\"data row2 col0\" >0.56</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row2_col1\" class=\"data row2 col1\" >nan</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row2_col2\" class=\"data row2 col2\" >0.55</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row2_col3\" class=\"data row2 col3\" >0.54</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529level0_row3\" class=\"row_heading level0 row3\" >nigeria</th>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row3_col0\" class=\"data row3 col0\" >nan</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row3_col1\" class=\"data row3 col1\" >0.30</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row3_col2\" class=\"data row3 col2\" >0.31</td>\n",
       "                        <td id=\"T_622c9060_9861_11e9_83d4_39d9ea9de529row3_col3\" class=\"data row3 col3\" >0.28</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fe4d42e54a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lsms_country_row_order = ['malawi', 'uganda', 'tanzania', 'nigeria']\n",
    "neal_ooc_lsms_df = neal_ooc(df=lsms_df, country_order=lsms_country_row_order)\n",
    "display(neal_ooc_lsms_df.style.apply(\n",
    "    background_gradient, cmap='coolwarm',\n",
    "    m=0.18, M=0.57,\n",
    "    axis=None).format('{:.2f}'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row0_col0 {\n",
       "            background-color:  #7396f5;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row0_col1 {\n",
       "            background-color:  #c4d5f3;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row0_col2 {\n",
       "            background-color:  #d9dce1;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row0_col3 {\n",
       "            background-color:  #f3c7b1;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row1_col0 {\n",
       "            background-color:  #c4d5f3;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row1_col1 {\n",
       "            background-color:  #e0dbd8;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row1_col2 {\n",
       "            background-color:  #f7a688;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row1_col3 {\n",
       "            background-color:  #d9dce1;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row2_col0 {\n",
       "            background-color:  #f7b89c;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row2_col1 {\n",
       "            background-color:  #bd1f2d;\n",
       "             color:  white;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row2_col2 {\n",
       "            background-color:  #f59c7d;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row2_col3 {\n",
       "            background-color:  #f29274;\n",
       "             color:  white;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row3_col0 {\n",
       "            background-color:  #f0cdbb;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row3_col1 {\n",
       "            background-color:  #a1c0ff;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row3_col2 {\n",
       "            background-color:  #7da0f9;\n",
       "        }    #T_631f65ba_9861_11e9_83d4_39d9ea9de529row3_col3 {\n",
       "            background-color:  #4257c9;\n",
       "             color:  white;\n",
       "        }</style><table id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529\" ><thead>    <tr>        <th class=\"index_name level0\" >train_on</th>        <th class=\"col_heading level0 col0\" >nigeria</th>        <th class=\"col_heading level0 col1\" >tanzania</th>        <th class=\"col_heading level0 col2\" >uganda</th>        <th class=\"col_heading level0 col3\" >malawi</th>    </tr>    <tr>        <th class=\"index_name level0\" >eval_on</th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529level0_row0\" class=\"row_heading level0 row0\" >malawi</th>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row0_col0\" class=\"data row0 col0\" >0.25</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row0_col1\" class=\"data row0 col1\" >0.34</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row0_col2\" class=\"data row0 col2\" >0.37</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row0_col3\" class=\"data row0 col3\" >0.42</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529level0_row1\" class=\"row_heading level0 row1\" >uganda</th>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row1_col0\" class=\"data row1 col0\" >0.34</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row1_col1\" class=\"data row1 col1\" >0.38</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row1_col2\" class=\"data row1 col2\" >0.46</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row1_col3\" class=\"data row1 col3\" >0.37</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529level0_row2\" class=\"row_heading level0 row2\" >tanzania</th>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row2_col0\" class=\"data row2 col0\" >0.44</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row2_col1\" class=\"data row2 col1\" >0.56</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row2_col2\" class=\"data row2 col2\" >0.47</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row2_col3\" class=\"data row2 col3\" >0.48</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529level0_row3\" class=\"row_heading level0 row3\" >nigeria</th>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row3_col0\" class=\"data row3 col0\" >0.41</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row3_col1\" class=\"data row3 col1\" >0.3</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row3_col2\" class=\"data row3 col2\" >0.26</td>\n",
       "                        <td id=\"T_631f65ba_9861_11e9_83d4_39d9ea9de529row3_col3\" class=\"data row3 col3\" >0.19</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fe454d84588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "neal_original_lsms_df = pd.DataFrame(data=[\n",
    "    [.25, .34, .37, .42],\n",
    "    [.34, .38, .46, .37],\n",
    "    [.44, .56, .47, .48],\n",
    "    [.41, .30, .26, .19]\n",
    "], index=lsms_country_row_order, columns=lsms_country_row_order[::-1])\n",
    "neal_original_lsms_df.index.name = 'eval_on'\n",
    "neal_original_lsms_df.columns.name = 'train_on'\n",
    "display(neal_original_lsms_df.style.apply(\n",
    "    background_gradient,\n",
    "    m=0.18, M=0.57,\n",
    "    cmap='coolwarm', axis=None))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row0_col0 {\n",
       "            background-color:  #92b4fe;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row0_col1 {\n",
       "            background-color:  #a5c3fe;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row0_col2 {\n",
       "            background-color:  #9bbcff;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row0_col3 {\n",
       "            background-color:  white;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row1_col0 {\n",
       "            background-color:  #f3c7b1;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row1_col1 {\n",
       "            background-color:  #ead5c9;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row1_col2 {\n",
       "            background-color:  white;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row1_col3 {\n",
       "            background-color:  #dedcdb;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row2_col0 {\n",
       "            background-color:  #c0282f;\n",
       "             color:  white;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row2_col1 {\n",
       "            background-color:  white;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row2_col2 {\n",
       "            background-color:  #cb3e38;\n",
       "             color:  white;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row2_col3 {\n",
       "            background-color:  #cf453c;\n",
       "             color:  white;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row3_col0 {\n",
       "            background-color:  white;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row3_col1 {\n",
       "            background-color:  #a5c3fe;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row3_col2 {\n",
       "            background-color:  #a6c4fe;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row3_col3 {\n",
       "            background-color:  #8fb1fe;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row4_col0 {\n",
       "            background-color:  #7396f5;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row4_col1 {\n",
       "            background-color:  #c4d5f3;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row4_col2 {\n",
       "            background-color:  #d9dce1;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row4_col3 {\n",
       "            background-color:  #f3c7b1;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row5_col0 {\n",
       "            background-color:  #c4d5f3;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row5_col1 {\n",
       "            background-color:  #e0dbd8;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row5_col2 {\n",
       "            background-color:  #f7a688;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row5_col3 {\n",
       "            background-color:  #d9dce1;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row6_col0 {\n",
       "            background-color:  #f7b89c;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row6_col1 {\n",
       "            background-color:  #bd1f2d;\n",
       "             color:  white;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row6_col2 {\n",
       "            background-color:  #f59c7d;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row6_col3 {\n",
       "            background-color:  #f29274;\n",
       "             color:  white;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row7_col0 {\n",
       "            background-color:  #f0cdbb;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row7_col1 {\n",
       "            background-color:  #a1c0ff;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row7_col2 {\n",
       "            background-color:  #7da0f9;\n",
       "        }    #T_6387e1d0_9861_11e9_83d4_39d9ea9de529row7_col3 {\n",
       "            background-color:  #4257c9;\n",
       "             color:  white;\n",
       "        }</style><table id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529\" ><thead>    <tr>        <th class=\"blank\" ></th>        <th class=\"index_name level0\" >train_on</th>        <th class=\"col_heading level0 col0\" >nigeria</th>        <th class=\"col_heading level0 col1\" >tanzania</th>        <th class=\"col_heading level0 col2\" >uganda</th>        <th class=\"col_heading level0 col3\" >malawi</th>    </tr>    <tr>        <th class=\"index_name level0\" ></th>        <th class=\"index_name level1\" >eval_on</th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529level0_row0\" class=\"row_heading level0 row0\" rowspan=4>KNN on scalar NL</th>\n",
       "                        <th id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529level1_row0\" class=\"row_heading level1 row0\" >malawi</th>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row0_col0\" class=\"data row0 col0\" >0.28</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row0_col1\" class=\"data row0 col1\" >0.30</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row0_col2\" class=\"data row0 col2\" >0.29</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row0_col3\" class=\"data row0 col3\" >nan</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529level1_row1\" class=\"row_heading level1 row1\" >uganda</th>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row1_col0\" class=\"data row1 col0\" >0.42</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row1_col1\" class=\"data row1 col1\" >0.40</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row1_col2\" class=\"data row1 col2\" >nan</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row1_col3\" class=\"data row1 col3\" >0.38</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529level1_row2\" class=\"row_heading level1 row2\" >tanzania</th>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row2_col0\" class=\"data row2 col0\" >0.56</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row2_col1\" class=\"data row2 col1\" >nan</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row2_col2\" class=\"data row2 col2\" >0.55</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row2_col3\" class=\"data row2 col3\" >0.54</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529level1_row3\" class=\"row_heading level1 row3\" >nigeria</th>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row3_col0\" class=\"data row3 col0\" >nan</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row3_col1\" class=\"data row3 col1\" >0.30</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row3_col2\" class=\"data row3 col2\" >0.31</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row3_col3\" class=\"data row3 col3\" >0.28</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529level0_row4\" class=\"row_heading level0 row4\" rowspan=4>Science paper</th>\n",
       "                        <th id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529level1_row4\" class=\"row_heading level1 row4\" >malawi</th>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row4_col0\" class=\"data row4 col0\" >0.25</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row4_col1\" class=\"data row4 col1\" >0.34</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row4_col2\" class=\"data row4 col2\" >0.37</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row4_col3\" class=\"data row4 col3\" >0.42</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529level1_row5\" class=\"row_heading level1 row5\" >uganda</th>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row5_col0\" class=\"data row5 col0\" >0.34</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row5_col1\" class=\"data row5 col1\" >0.38</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row5_col2\" class=\"data row5 col2\" >0.46</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row5_col3\" class=\"data row5 col3\" >0.37</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529level1_row6\" class=\"row_heading level1 row6\" >tanzania</th>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row6_col0\" class=\"data row6 col0\" >0.44</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row6_col1\" class=\"data row6 col1\" >0.56</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row6_col2\" class=\"data row6 col2\" >0.47</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row6_col3\" class=\"data row6 col3\" >0.48</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529level1_row7\" class=\"row_heading level1 row7\" >nigeria</th>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row7_col0\" class=\"data row7 col0\" >0.41</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row7_col1\" class=\"data row7 col1\" >0.30</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row7_col2\" class=\"data row7 col2\" >0.26</td>\n",
       "                        <td id=\"T_6387e1d0_9861_11e9_83d4_39d9ea9de529row7_col3\" class=\"data row7 col3\" >0.19</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fe454d88390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "concat_df = pd.concat(\n",
    "    [neal_ooc_lsms_df, neal_original_lsms_df],\n",
    "    keys=['KNN on scalar NL', 'Science paper'])\n",
    "display(concat_df.style.apply(\n",
    "    background_gradient, cmap='coolwarm',\n",
    "    m=0.18, M=0.57,\n",
    "    axis=None).format('{:.2f}'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Neal NL and Labels vs. new NL and Labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Neal Label vs. Label: r^2 = 0.5900936204254654\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": [
    "sns.scatterplot(data=df, x='neal_label', y='label', hue='country')\n",
    "r2 = calc_score(df['neal_label'], df['label'], metric='r2')\n",
    "print('Neal Label vs. Label: r^2 =', r2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Neal NL vs. NL: r^2 = 0.7804754226120915\n",
      "*Note: NL for Nigeria (2013) is VIIRS, while all others are DMSP\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": [
    "sns.scatterplot(data=df, x='neal_nl', y='nl', hue='country')\n",
    "r2 = calc_score(df['neal_nl'], df['nl'], metric='r2')\n",
    "print('Neal NL vs. NL: r^2 =', r2)\n",
    "print('*Note: NL for Nigeria (2013) is VIIRS, while all others are DMSP')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(1, 4, figsize=(10, 3))\n",
    "i = 0\n",
    "for label_col in ['label', 'neal_label']:\n",
    "    for nl_col in ['neal_nl', 'nl']:\n",
    "        r2 = calc_score(df[label_col], df[nl_col], metric='r2')\n",
    "        df.plot.scatter(x=label_col, y=nl_col, ax=axs[i], label=f'$r^2 = {r2:.3f}$')\n",
    "        axs[i].legend()\n",
    "        i += 1\n",
    "fig.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def knn_nl(labels, years, nls_center, train_idxs, test_idxs, k=5):\n",
    "    '''\n",
    "    Args\n",
    "    - labels: np.array, shape [N]\n",
    "    - years: np.array, shape [N]\n",
    "    - nls_center: np.array, shape [N]\n",
    "    - splits: dict, split (str) => np.array\n",
    "    - k: int, k for KNN\n",
    "    '''\n",
    "    train_nls = nls_center[train_idxs]\n",
    "\n",
    "    # test\n",
    "    test_nls = nls_center[test_idxs]\n",
    "    test_labels = labels[test_idxs]\n",
    "    preds = np.zeros_like(test_labels)\n",
    "\n",
    "    knn_dmsp = sklearn.neighbors.KNeighborsRegressor(k)\n",
    "    knn_viirs = sklearn.neighbors.KNeighborsRegressor(k)\n",
    "\n",
    "    dmsp_mask = years < 2012\n",
    "    viirs_mask = ~dmsp_mask\n",
    "\n",
    "    for mask, model in [(dmsp_mask, knn_dmsp), (viirs_mask, knn_viirs)]:\n",
    "        train_mask = mask[train_idxs]\n",
    "        test_mask = mask[test_idxs]\n",
    "\n",
    "        if np.sum(test_mask) == 0:\n",
    "            continue\n",
    "\n",
    "        # assign each unique NL value the same label\n",
    "        u, indices = np.unique(train_nls[train_mask], return_inverse=True)\n",
    "        mean_labels = np.zeros_like(u)\n",
    "        for i, nl in enumerate(u):\n",
    "            mean_labels[i] = np.mean(labels[nls_center == nl])\n",
    "\n",
    "        model.fit(X=train_nls[train_mask].reshape(-1, 1), y=mean_labels[indices])\n",
    "        preds[test_mask] = model.predict(X=test_nls[test_mask].reshape(-1, 1))\n",
    "\n",
    "    return preds, knn_dmsp, knn_viirs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def knn_cv_wrapper(labels, years, nls, folds, keep=1.0, seed=None, savepath=None, verbose=False):\n",
    "    '''\n",
    "    Args\n",
    "    - labels: np.array, shape [N]\n",
    "    - years: np.array, shape [N]\n",
    "    - nls: np.array, shape [N]\n",
    "    - folds: dict, folds[f][split] is a np.array of indices\n",
    "    - keep: float, fraction of training and validation data to use\n",
    "    - seed: int, random seed\n",
    "    - savepath: str, path to save predictions\n",
    "    - verbose: bool, whether to print out validation performance on each fold\n",
    "\n",
    "    Returns\n",
    "    - best_models: tuple of sklearn.neighbors.KNeighborsRegressor\n",
    "        best models from the last fold\n",
    "    - test_preds: np.array, shape [N]\n",
    "    - ks: np.array, shape [K]\n",
    "    - r2s: np.array, shape [K]\n",
    "    - R2s: np.array, shape [K]\n",
    "    - savepath: str, path to save npz file\n",
    "    '''\n",
    "    N = len(labels)\n",
    "    assert len(years) == N\n",
    "    assert len(nls) == N\n",
    "    if savepath is not None:\n",
    "        assert not os.path.exists(savepath)\n",
    "\n",
    "    test_preds = np.zeros_like(labels)\n",
    "    ks = np.asarray([1, 2, 5, 10, 30, 50, 100, 200, 500, 1000])\n",
    "    fold_keys = sorted(folds.keys())\n",
    "    r2s = np.ones([len(fold_keys), len(ks)]) * (-np.inf)\n",
    "    R2s = np.ones([len(fold_keys), len(ks)]) * (-np.inf)\n",
    "\n",
    "    dmsp_mask = years < 2012\n",
    "    viirs_mask = ~dmsp_mask\n",
    "\n",
    "    for i, f in enumerate(fold_keys):\n",
    "        if verbose:\n",
    "            print(f'Runing fold {f}...')\n",
    "        test_idxs = folds[f]['test']\n",
    "        best_models = None\n",
    "\n",
    "        for j, k in enumerate(ks):\n",
    "            train_idxs = folds[f]['train']\n",
    "            val_idxs = folds[f]['val']\n",
    "\n",
    "            num_train = int(len(train_idxs) * keep)\n",
    "            num_val = int(len(val_idxs) * keep)\n",
    "            if num_train < k or num_val < k:\n",
    "                print('num_train:', num_train, 'num_val:', num_train, 'k:', k)\n",
    "                continue\n",
    "\n",
    "            if keep < 1:\n",
    "                if seed is not None:\n",
    "                    np.random.seed(seed)\n",
    "                train_idxs = np.random.choice(\n",
    "                    train_idxs, size=num_train, replace=False)\n",
    "                val_tfrecord_paths = np.random.choice(\n",
    "                    val_idxs, size=num_val, replace=False)\n",
    "\n",
    "            val_preds, knn_dmsp, knn_viirs = knn_nl(\n",
    "                labels, years, nls, train_idxs=train_idxs,\n",
    "                test_idxs=val_idxs, k=k)\n",
    "            r2 = calc_score(labels=labels[val_idxs], preds=val_preds, metric='r2')\n",
    "            R2 = calc_score(labels=labels[val_idxs], preds=val_preds, metric='R2')\n",
    "            if R2 > np.max(R2s[i]):\n",
    "                best_models = (knn_dmsp, knn_viirs)\n",
    "            r2s[i, j], R2s[i, j] = r2, R2\n",
    "            if verbose:\n",
    "                print(f'fold={f}, k={k:4d}, val_r2={r2:0.4f}, val_R2={R2:0.4f}')\n",
    "\n",
    "        for i, mask in enumerate([dmsp_mask, viirs_mask]):\n",
    "            idxs = np.nonzero(mask)[0]\n",
    "            if len(idxs) <= 0:\n",
    "                continue\n",
    "            test_preds[idxs] = best_models[i].predict(nls[idxs].reshape(-1, 1))\n",
    "\n",
    "    evaluate(labels=labels, preds=test_preds, do_print=True, title='test preds')\n",
    "\n",
    "    if savepath is not None:\n",
    "        os.makedirs(os.path.dirname(savepath), exist_ok=True)\n",
    "        np.savez_compressed(savepath, test_preds=test_preds, labels=labels)\n",
    "        \n",
    "    return best_models, test_preds, ks, r2s, R2s\n",
    "\n",
    "def plot_knn_val_performance(nl_result, labels, title):\n",
    "    '''\n",
    "    Args\n",
    "    - nl_result: tuple (models, ks, r2s, R2s), output of knn_cv_wrapper\n",
    "    - labels: np.array\n",
    "    - title: str\n",
    "    '''\n",
    "    _, test_preds, ks, r2s, R2s = nl_result\n",
    "    fig, axs = plt.subplots(1, 2, sharey=True, figsize=(6, 3))\n",
    "    for i, f in enumerate(FOLDS):\n",
    "        print(f'fold {f}, chosen k:', ks[np.argmax(R2s[i])])\n",
    "        axs[0].plot(ks, r2s[i], label=f'fold={f}, $r^2$')\n",
    "        axs[1].plot(ks, R2s[i], label=f'fold={f}, $R^2$')\n",
    "    for ax in axs:\n",
    "        ax.legend()\n",
    "        ax.grid()\n",
    "        ax.set(xlabel='k', xscale='log')\n",
    "    axs[0].set(ylabel='score')\n",
    "    fig.suptitle(f'{title} validation KNN performance vs. K', y=1.02)\n",
    "    fig.tight_layout()\n",
    "    plt.show()\n",
    "\n",
    "def plot_knn_test(nls_center_result, nls_mean_result, labels,\n",
    "                  nl0_center_mask, nl0_mean_mask, title):\n",
    "    '''\n",
    "    Args\n",
    "    - *_result: tuple (models, ks, r2s, R2s), output of knn_cv_wrapper\n",
    "    - labels: np.array, shape [N]\n",
    "    - *_mask: np.array, shape [N]\n",
    "    - title: str\n",
    "    '''\n",
    "    fig, axs = plt.subplots(1, 2, sharey=True, figsize=(6, 3))\n",
    "    results = [nls_center_result, nls_mean_result]\n",
    "    nl0_masks = [nl0_center_mask, nl0_mean_mask]\n",
    "    for result, nl0_mask, ax in zip(results, nl0_masks, axs):\n",
    "        _, test_preds, ks, r2s, R2s = result\n",
    "        scatter_preds(labels, test_preds, ax=ax)\n",
    "        ax.scatter(x=labels[nl0_mask], y=test_preds[nl0_mask], c='red', s=2, label='NL = 0')\n",
    "        ax.legend()\n",
    "    axs[0].set_title(f'KNN on NL Center ({title})')\n",
    "    axs[1].set_title(f'KNN on NL Mean ({title})')\n",
    "    fig.tight_layout()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 'A': {\n",
    "#     'train': np.array([1, 10, 13, ...]),\n",
    "#     ...\n",
    "# }\n",
    "ooc_folds = {\n",
    "    f: {split: [] for split in SPLITS}\n",
    "    for f in FOLDS\n",
    "}\n",
    "\n",
    "for i, f in enumerate(FOLDS):\n",
    "    ooc_folds[f] = {}\n",
    "    test_country = DMSP_COUNTRIES[i]\n",
    "    val_country = DMSP_COUNTRIES[(i+1)%4]\n",
    "    train_countries = set(DMSP_COUNTRIES) - {test_country, val_country}\n",
    "    ooc_folds[f]['test'] = np.nonzero((df['country'] == test_country).values)[0]\n",
    "    ooc_folds[f]['val'] = np.nonzero((df['country'] == val_country).values)[0]\n",
    "    ooc_folds[f]['train'] = np.nonzero(df['country'].isin(train_countries).values)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "DMSP0 = -MEANS['DMSP'] / STD_DEVS['DMSP']\n",
    "VIIRS0 = -MEANS['VIIRS'] / STD_DEVS['VIIRS']\n",
    "nl0_center_mask = ((nls_center == DMSP0) & dmsp_mask) | ((nls_center == VIIRS0) & viirs_mask)\n",
    "nl0_mean_mask = ((nls_mean == DMSP0) & dmsp_mask) | ((nls_mean == VIIRS0) & viirs_mask)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "country_indices = {\n",
    "    country: np.nonzero((df['country'] == country).values)[0]\n",
    "    for country in COUNTRIES\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Neal OOC setup\n",
    "Choose k=100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for survey in surveys:\n",
    "    country, year = survey.split('_dhs_')\n",
    "    year = int(year)\n",
    "\n",
    "    folder = os.path.join(NEAL_DIR, survey, 'data')\n",
    "    image_counts = np.load(os.path.join(folder, 'cluster_image_counts.npy'))\n",
    "    mask = (image_counts >= 10)\n",
    "\n",
    "    clocs = np.load(os.path.join(folder, 'cluster_locs.npy'))\n",
    "    clocs = clocs[mask]\n",
    "\n",
    "    neal_nls = np.load(os.path.join(folder, 'cluster_nightlights.npy'))\n",
    "    neal_nls = neal_nls[mask]\n",
    "\n",
    "    neal_labels = np.load(os.path.join(folder, 'cluster_assets.npy'))\n",
    "    neal_labels = neal_labels[mask]\n",
    "\n",
    "    indices = [\n",
    "        loc_to_index[(lat, lon)]\n",
    "        for lat, lon in clocs.astype(np.float32)\n",
    "    ]\n",
    "    \n",
    "    cdf = pd.DataFrame(data={\n",
    "        'lat': clocs[:, 0],\n",
    "        'lon': clocs[:, 1],\n",
    "        'country': country,\n",
    "        'year': year,\n",
    "        'neal_nl': neal_nls,\n",
    "        'neal_label': neal_labels,\n",
    "        'nl': nls_mean[indices],\n",
    "        'label': labels[indices]\n",
    "    })\n",
    "    df = pd.concat([df, cdf])\n",
    "df.reset_index(inplace=True, drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 283,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row0_col0 {\n",
       "            background-color:  #bd1f2d;\n",
       "             color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row0_col1 {\n",
       "            background-color:  #c53334;\n",
       "             color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row0_col2 {\n",
       "            background-color:  #bb1b2c;\n",
       "             color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row0_col3 {\n",
       "            background-color:  #b40426;\n",
       "             color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row0_col4 {\n",
       "            background-color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row1_col0 {\n",
       "            background-color:  #f7af91;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row1_col1 {\n",
       "            background-color:  #f7ba9f;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row1_col2 {\n",
       "            background-color:  #f7b89c;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row1_col3 {\n",
       "            background-color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row1_col4 {\n",
       "            background-color:  #f1ccb8;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row2_col0 {\n",
       "            background-color:  #ec8165;\n",
       "             color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row2_col1 {\n",
       "            background-color:  #e9785d;\n",
       "             color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row2_col2 {\n",
       "            background-color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row2_col3 {\n",
       "            background-color:  #f29072;\n",
       "             color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row2_col4 {\n",
       "            background-color:  #f4c5ad;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row3_col0 {\n",
       "            background-color:  #dd5f4b;\n",
       "             color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row3_col1 {\n",
       "            background-color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row3_col2 {\n",
       "            background-color:  #f5a081;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row3_col3 {\n",
       "            background-color:  #e57058;\n",
       "             color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row3_col4 {\n",
       "            background-color:  #f2c9b4;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row4_col0 {\n",
       "            background-color:  white;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row4_col1 {\n",
       "            background-color:  #f39577;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row4_col2 {\n",
       "            background-color:  #f6bea4;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row4_col3 {\n",
       "            background-color:  #f59c7d;\n",
       "        }    #T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row4_col4 {\n",
       "            background-color:  #ead5c9;\n",
       "        }</style><table id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7\" ><thead>    <tr>        <th class=\"index_name level0\" >train_on</th>        <th class=\"col_heading level0 col0\" >nigeria</th>        <th class=\"col_heading level0 col1\" >tanzania</th>        <th class=\"col_heading level0 col2\" >uganda</th>        <th class=\"col_heading level0 col3\" >malawi</th>        <th class=\"col_heading level0 col4\" >rwanda</th>    </tr>    <tr>        <th class=\"index_name level0\" >eval_on</th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7level0_row0\" class=\"row_heading level0 row0\" >rwanda</th>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row0_col0\" class=\"data row0 col0\" >0.78</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row0_col1\" class=\"data row0 col1\" >0.76</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row0_col2\" class=\"data row0 col2\" >0.78</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row0_col3\" class=\"data row0 col3\" >0.79</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row0_col4\" class=\"data row0 col4\" >nan</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7level0_row1\" class=\"row_heading level0 row1\" >malawi</th>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row1_col0\" class=\"data row1 col0\" >0.62</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row1_col1\" class=\"data row1 col1\" >0.60</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row1_col2\" class=\"data row1 col2\" >0.61</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row1_col3\" class=\"data row1 col3\" >nan</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row1_col4\" class=\"data row1 col4\" >0.57</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7level0_row2\" class=\"row_heading level0 row2\" >uganda</th>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row2_col0\" class=\"data row2 col0\" >0.68</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row2_col1\" class=\"data row2 col1\" >0.69</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row2_col2\" class=\"data row2 col2\" >nan</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row2_col3\" class=\"data row2 col3\" >0.67</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row2_col4\" class=\"data row2 col4\" >0.58</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7level0_row3\" class=\"row_heading level0 row3\" >tanzania</th>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row3_col0\" class=\"data row3 col0\" >0.72</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row3_col1\" class=\"data row3 col1\" >nan</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row3_col2\" class=\"data row3 col2\" >0.64</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row3_col3\" class=\"data row3 col3\" >0.70</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row3_col4\" class=\"data row3 col4\" >0.57</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7level0_row4\" class=\"row_heading level0 row4\" >nigeria</th>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row4_col0\" class=\"data row4 col0\" >nan</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row4_col1\" class=\"data row4 col1\" >0.66</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row4_col2\" class=\"data row4 col2\" >0.60</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row4_col3\" class=\"data row4 col3\" >0.65</td>\n",
       "                        <td id=\"T_45e13f0a_8786_11e9_beba_83cd8f8b7bd7row4_col4\" class=\"data row4 col4\" >0.54</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fcad4a20048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dhs_country_row_order = ['rwanda', 'malawi', 'uganda', 'tanzania', 'nigeria']\n",
    "neal_ooc_df = neal_ooc(df=df, country_order=dhs_country_row_order)\n",
    "display(neal_ooc_df.style.apply(\n",
    "    background_gradient, cmap='coolwarm',\n",
    "    m=0.24, M=0.79,\n",
    "    axis=None).format('{:.2f}'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 285,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row0_col0 {\n",
       "            background-color:  #ecd3c5;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row0_col1 {\n",
       "            background-color:  #ee8468;\n",
       "             color:  white;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row0_col2 {\n",
       "            background-color:  #f39475;\n",
       "             color:  white;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row0_col3 {\n",
       "            background-color:  #f5c1a9;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row0_col4 {\n",
       "            background-color:  #cd423b;\n",
       "             color:  white;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row1_col0 {\n",
       "            background-color:  #9bbcff;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row1_col1 {\n",
       "            background-color:  #a7c5fe;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row1_col2 {\n",
       "            background-color:  #aec9fc;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row1_col3 {\n",
       "            background-color:  #ecd3c5;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row1_col4 {\n",
       "            background-color:  #a7c5fe;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row2_col0 {\n",
       "            background-color:  #efcfbf;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row2_col1 {\n",
       "            background-color:  #f7b599;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row2_col2 {\n",
       "            background-color:  #f39475;\n",
       "             color:  white;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row2_col3 {\n",
       "            background-color:  #ecd3c5;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row2_col4 {\n",
       "            background-color:  #efcfbf;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row3_col0 {\n",
       "            background-color:  #aec9fc;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row3_col1 {\n",
       "            background-color:  #f4c6af;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row3_col2 {\n",
       "            background-color:  #dadce0;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row3_col3 {\n",
       "            background-color:  #9bbcff;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row3_col4 {\n",
       "            background-color:  #c6d6f1;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row4_col0 {\n",
       "            background-color:  #ea7b60;\n",
       "             color:  white;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row4_col1 {\n",
       "            background-color:  #d6dce4;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row4_col2 {\n",
       "            background-color:  #c0d4f5;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row4_col3 {\n",
       "            background-color:  #3b4cc0;\n",
       "             color:  white;\n",
       "        }    #T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row4_col4 {\n",
       "            background-color:  #94b6ff;\n",
       "        }</style><table id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7\" ><thead>    <tr>        <th class=\"index_name level0\" >train_on</th>        <th class=\"col_heading level0 col0\" >nigeria</th>        <th class=\"col_heading level0 col1\" >tanzania</th>        <th class=\"col_heading level0 col2\" >uganda</th>        <th class=\"col_heading level0 col3\" >malawi</th>        <th class=\"col_heading level0 col4\" >rwanda</th>    </tr>    <tr>        <th class=\"index_name level0\" >eval_on</th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7level0_row0\" class=\"row_heading level0 row0\" >rwanda</th>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row0_col0\" class=\"data row0 col0\" >0.55</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row0_col1\" class=\"data row0 col1\" >0.68</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row0_col2\" class=\"data row0 col2\" >0.66</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row0_col3\" class=\"data row0 col3\" >0.59</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row0_col4\" class=\"data row0 col4\" >0.75</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7level0_row1\" class=\"row_heading level0 row1\" >malawi</th>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row1_col0\" class=\"data row1 col0\" >0.4</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row1_col1\" class=\"data row1 col1\" >0.42</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row1_col2\" class=\"data row1 col2\" >0.43</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row1_col3\" class=\"data row1 col3\" >0.55</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row1_col4\" class=\"data row1 col4\" >0.42</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7level0_row2\" class=\"row_heading level0 row2\" >uganda</th>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row2_col0\" class=\"data row2 col0\" >0.56</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row2_col1\" class=\"data row2 col1\" >0.61</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row2_col2\" class=\"data row2 col2\" >0.66</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row2_col3\" class=\"data row2 col3\" >0.55</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row2_col4\" class=\"data row2 col4\" >0.56</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7level0_row3\" class=\"row_heading level0 row3\" >tanzania</th>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row3_col0\" class=\"data row3 col0\" >0.43</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row3_col1\" class=\"data row3 col1\" >0.58</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row3_col2\" class=\"data row3 col2\" >0.51</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row3_col3\" class=\"data row3 col3\" >0.4</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row3_col4\" class=\"data row3 col4\" >0.47</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7level0_row4\" class=\"row_heading level0 row4\" >nigeria</th>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row4_col0\" class=\"data row4 col0\" >0.69</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row4_col1\" class=\"data row4 col1\" >0.5</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row4_col2\" class=\"data row4 col2\" >0.46</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row4_col3\" class=\"data row4 col3\" >0.24</td>\n",
       "                        <td id=\"T_5c28d322_8786_11e9_beba_83cd8f8b7bd7row4_col4\" class=\"data row4 col4\" >0.39</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fcad4bb02e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "neal_original_df = pd.DataFrame(data=[\n",
    "    [.55, .68, .66, .59, .75],\n",
    "    [.40, .42, .43, .55, .42],\n",
    "    [.56, .61, .66, .55, .56],\n",
    "    [.43, .58, .51, .40, .47],\n",
    "    [.69, .50, .46, .24, .39],\n",
    "], index=dhs_country_row_order, columns=dhs_country_row_order[::-1])\n",
    "neal_original_df.index.name = 'eval_on'\n",
    "neal_original_df.columns.name = 'train_on'\n",
    "display(neal_original_df.style.apply(\n",
    "    background_gradient,\n",
    "    m=0.24, M=0.79,\n",
    "    cmap='coolwarm', axis=None))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 286,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row0_col0 {\n",
       "            background-color:  #bd1f2d;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row0_col1 {\n",
       "            background-color:  #c53334;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row0_col2 {\n",
       "            background-color:  #bb1b2c;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row0_col3 {\n",
       "            background-color:  #b40426;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row0_col4 {\n",
       "            background-color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row1_col0 {\n",
       "            background-color:  #f7af91;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row1_col1 {\n",
       "            background-color:  #f7ba9f;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row1_col2 {\n",
       "            background-color:  #f7b89c;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row1_col3 {\n",
       "            background-color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row1_col4 {\n",
       "            background-color:  #f1ccb8;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row2_col0 {\n",
       "            background-color:  #ec8165;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row2_col1 {\n",
       "            background-color:  #e9785d;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row2_col2 {\n",
       "            background-color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row2_col3 {\n",
       "            background-color:  #f29072;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row2_col4 {\n",
       "            background-color:  #f4c5ad;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row3_col0 {\n",
       "            background-color:  #dd5f4b;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row3_col1 {\n",
       "            background-color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row3_col2 {\n",
       "            background-color:  #f5a081;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row3_col3 {\n",
       "            background-color:  #e57058;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row3_col4 {\n",
       "            background-color:  #f2c9b4;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row4_col0 {\n",
       "            background-color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row4_col1 {\n",
       "            background-color:  #f39577;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row4_col2 {\n",
       "            background-color:  #f6bea4;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row4_col3 {\n",
       "            background-color:  #f59c7d;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row4_col4 {\n",
       "            background-color:  #ead5c9;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row5_col0 {\n",
       "            background-color:  #ecd3c5;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row5_col1 {\n",
       "            background-color:  #ee8468;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row5_col2 {\n",
       "            background-color:  #f39475;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row5_col3 {\n",
       "            background-color:  #f5c1a9;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row5_col4 {\n",
       "            background-color:  #cd423b;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row6_col0 {\n",
       "            background-color:  #9bbcff;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row6_col1 {\n",
       "            background-color:  #a7c5fe;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row6_col2 {\n",
       "            background-color:  #aec9fc;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row6_col3 {\n",
       "            background-color:  #ecd3c5;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row6_col4 {\n",
       "            background-color:  #a7c5fe;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row7_col0 {\n",
       "            background-color:  #efcfbf;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row7_col1 {\n",
       "            background-color:  #f7b599;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row7_col2 {\n",
       "            background-color:  #f39475;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row7_col3 {\n",
       "            background-color:  #ecd3c5;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row7_col4 {\n",
       "            background-color:  #efcfbf;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row8_col0 {\n",
       "            background-color:  #aec9fc;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row8_col1 {\n",
       "            background-color:  #f4c6af;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row8_col2 {\n",
       "            background-color:  #dadce0;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row8_col3 {\n",
       "            background-color:  #9bbcff;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row8_col4 {\n",
       "            background-color:  #c6d6f1;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row9_col0 {\n",
       "            background-color:  #ea7b60;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row9_col1 {\n",
       "            background-color:  #d6dce4;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row9_col2 {\n",
       "            background-color:  #c0d4f5;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row9_col3 {\n",
       "            background-color:  #3b4cc0;\n",
       "             color:  white;\n",
       "        }    #T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row9_col4 {\n",
       "            background-color:  #94b6ff;\n",
       "        }</style><table id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7\" ><thead>    <tr>        <th class=\"blank\" ></th>        <th class=\"index_name level0\" >train_on</th>        <th class=\"col_heading level0 col0\" >nigeria</th>        <th class=\"col_heading level0 col1\" >tanzania</th>        <th class=\"col_heading level0 col2\" >uganda</th>        <th class=\"col_heading level0 col3\" >malawi</th>        <th class=\"col_heading level0 col4\" >rwanda</th>    </tr>    <tr>        <th class=\"index_name level0\" ></th>        <th class=\"index_name level1\" >eval_on</th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level0_row0\" class=\"row_heading level0 row0\" rowspan=5>KNN on scalar NL</th>\n",
       "                        <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level1_row0\" class=\"row_heading level1 row0\" >rwanda</th>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row0_col0\" class=\"data row0 col0\" >0.78</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row0_col1\" class=\"data row0 col1\" >0.76</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row0_col2\" class=\"data row0 col2\" >0.78</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row0_col3\" class=\"data row0 col3\" >0.79</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row0_col4\" class=\"data row0 col4\" >nan</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level1_row1\" class=\"row_heading level1 row1\" >malawi</th>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row1_col0\" class=\"data row1 col0\" >0.62</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row1_col1\" class=\"data row1 col1\" >0.60</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row1_col2\" class=\"data row1 col2\" >0.61</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row1_col3\" class=\"data row1 col3\" >nan</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row1_col4\" class=\"data row1 col4\" >0.57</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level1_row2\" class=\"row_heading level1 row2\" >uganda</th>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row2_col0\" class=\"data row2 col0\" >0.68</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row2_col1\" class=\"data row2 col1\" >0.69</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row2_col2\" class=\"data row2 col2\" >nan</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row2_col3\" class=\"data row2 col3\" >0.67</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row2_col4\" class=\"data row2 col4\" >0.58</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level1_row3\" class=\"row_heading level1 row3\" >tanzania</th>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row3_col0\" class=\"data row3 col0\" >0.72</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row3_col1\" class=\"data row3 col1\" >nan</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row3_col2\" class=\"data row3 col2\" >0.64</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row3_col3\" class=\"data row3 col3\" >0.70</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row3_col4\" class=\"data row3 col4\" >0.57</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level1_row4\" class=\"row_heading level1 row4\" >nigeria</th>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row4_col0\" class=\"data row4 col0\" >nan</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row4_col1\" class=\"data row4 col1\" >0.66</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row4_col2\" class=\"data row4 col2\" >0.60</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row4_col3\" class=\"data row4 col3\" >0.65</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row4_col4\" class=\"data row4 col4\" >0.54</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level0_row5\" class=\"row_heading level0 row5\" rowspan=5>Science paper</th>\n",
       "                        <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level1_row5\" class=\"row_heading level1 row5\" >rwanda</th>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row5_col0\" class=\"data row5 col0\" >0.55</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row5_col1\" class=\"data row5 col1\" >0.68</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row5_col2\" class=\"data row5 col2\" >0.66</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row5_col3\" class=\"data row5 col3\" >0.59</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row5_col4\" class=\"data row5 col4\" >0.75</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level1_row6\" class=\"row_heading level1 row6\" >malawi</th>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row6_col0\" class=\"data row6 col0\" >0.40</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row6_col1\" class=\"data row6 col1\" >0.42</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row6_col2\" class=\"data row6 col2\" >0.43</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row6_col3\" class=\"data row6 col3\" >0.55</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row6_col4\" class=\"data row6 col4\" >0.42</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level1_row7\" class=\"row_heading level1 row7\" >uganda</th>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row7_col0\" class=\"data row7 col0\" >0.56</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row7_col1\" class=\"data row7 col1\" >0.61</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row7_col2\" class=\"data row7 col2\" >0.66</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row7_col3\" class=\"data row7 col3\" >0.55</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row7_col4\" class=\"data row7 col4\" >0.56</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level1_row8\" class=\"row_heading level1 row8\" >tanzania</th>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row8_col0\" class=\"data row8 col0\" >0.43</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row8_col1\" class=\"data row8 col1\" >0.58</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row8_col2\" class=\"data row8 col2\" >0.51</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row8_col3\" class=\"data row8 col3\" >0.40</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row8_col4\" class=\"data row8 col4\" >0.47</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <th id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7level1_row9\" class=\"row_heading level1 row9\" >nigeria</th>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row9_col0\" class=\"data row9 col0\" >0.69</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row9_col1\" class=\"data row9 col1\" >0.50</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row9_col2\" class=\"data row9 col2\" >0.46</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row9_col3\" class=\"data row9 col3\" >0.24</td>\n",
       "                        <td id=\"T_5f3b5f1c_8786_11e9_beba_83cd8f8b7bd7row9_col4\" class=\"data row9 col4\" >0.39</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fcad4b720b8>"
      ]
     },
     "execution_count": 286,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "concat_df = pd.concat(\n",
    "    [neal_ooc_df, neal_original_df],\n",
    "    keys=['KNN on scalar NL', 'Science paper'])\n",
    "display(concat_df.style.apply(\n",
    "    background_gradient, cmap='coolwarm',\n",
    "    m=0.24, M=0.79,\n",
    "    axis=None).format('{:.2f}'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Label vs. NL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num_train: 848 num_val: 848 k: 500\n",
      "num_train: 848 num_val: 848 k: 1000\n",
      "num_train: 1220 num_val: 1220 k: 500\n",
      "num_train: 1220 num_val: 1220 k: 1000\n",
      "num_train: 1319 num_val: 1319 k: 500\n",
      "num_train: 1319 num_val: 1319 k: 1000\n",
      "num_train: 947 num_val: 947 k: 1000\n",
      "test preds\t- r^2: 0.641, R^2: 0.581, mse: 0.133, rank: 0.650\n",
      "malawi\t- r^2: 0.559, R^2: 0.547, mse: 0.104, rank: 0.590\n",
      "rwanda\t- r^2: 0.717, R^2: 0.664, mse: 0.083, rank: 0.555\n",
      "tanzania\t- r^2: 0.633, R^2: 0.512, mse: 0.206, rank: 0.759\n",
      "uganda\t- r^2: 0.723, R^2: 0.593, mse: 0.172, rank: 0.752\n"
     ]
    }
   ],
   "source": [
    "nls_mean_result = knn_cv_wrapper(\n",
    "    labels=df['label'].values, years=df['year'].values, nls=df['nl'].values, folds=ooc_folds,\n",
    "    savepath=None)\n",
    "test_preds = nls_mean_result[1]\n",
    "for country in COUNTRIES:\n",
    "    indices = country_indices[country]\n",
    "    evaluate(labels=df.loc[indices, 'label'], preds=test_preds[indices], do_print=True, title=country)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fold A, chosen k: 100\n",
      "fold B, chosen k: 30\n",
      "fold C, chosen k: 30\n",
      "fold D, chosen k: 200\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_knn_val_performance(nls_mean_result, df['label'].values, title='NL mean')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7fab0c8b0390>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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.scatter(df['label'], test_preds, c=df['country'].map(COUNTRIES.index), cmap='tab10')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Neal Label vs. Neal NL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num_train: 848 num_val: 848 k: 500\n",
      "num_train: 848 num_val: 848 k: 1000\n",
      "num_train: 1220 num_val: 1220 k: 500\n",
      "num_train: 1220 num_val: 1220 k: 1000\n",
      "num_train: 1319 num_val: 1319 k: 500\n",
      "num_train: 1319 num_val: 1319 k: 1000\n",
      "num_train: 947 num_val: 947 k: 1000\n",
      "test preds\t- r^2: 0.673, R^2: 0.671, mse: 0.232, rank: 0.636\n",
      "malawi\t- r^2: 0.616, R^2: 0.601, mse: 0.265, rank: 0.579\n",
      "rwanda\t- r^2: 0.793, R^2: 0.764, mse: 0.164, rank: 0.485\n",
      "tanzania\t- r^2: 0.701, R^2: 0.700, mse: 0.212, rank: 0.752\n",
      "uganda\t- r^2: 0.665, R^2: 0.656, mse: 0.270, rank: 0.712\n"
     ]
    }
   ],
   "source": [
    "nls_mean_result = knn_cv_wrapper(\n",
    "    labels=df['neal_label'].values, years=df['year'].values, nls=df['neal_nl'].values, folds=ooc_folds,\n",
    "    savepath=None)\n",
    "test_preds = nls_mean_result[1]\n",
    "for country in COUNTRIES:\n",
    "    indices = country_indices[country]\n",
    "    evaluate(labels=df.loc[indices, 'neal_label'], preds=test_preds[indices], do_print=True, title=country)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fold A, chosen k: 30\n",
      "fold B, chosen k: 100\n",
      "fold C, chosen k: 200\n",
      "fold D, chosen k: 100\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_knn_val_performance(nls_mean_result, labels, title='NL mean')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7fab0d325cc0>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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.scatter(df['neal_label'], test_preds, c=df['country'].map(COUNTRIES.index), cmap='tab10')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Neal Label vs. NL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num_train: 848 num_val: 848 k: 500\n",
      "num_train: 848 num_val: 848 k: 1000\n",
      "num_train: 1220 num_val: 1220 k: 500\n",
      "num_train: 1220 num_val: 1220 k: 1000\n",
      "num_train: 1319 num_val: 1319 k: 500\n",
      "num_train: 1319 num_val: 1319 k: 1000\n",
      "num_train: 947 num_val: 947 k: 1000\n",
      "test preds\t- r^2: 0.662, R^2: 0.657, mse: 0.242, rank: 0.655\n",
      "malawi\t- r^2: 0.592, R^2: 0.581, mse: 0.278, rank: 0.600\n",
      "rwanda\t- r^2: 0.775, R^2: 0.769, mse: 0.161, rank: 0.498\n",
      "tanzania\t- r^2: 0.699, R^2: 0.691, mse: 0.219, rank: 0.757\n",
      "uganda\t- r^2: 0.682, R^2: 0.629, mse: 0.291, rank: 0.736\n"
     ]
    }
   ],
   "source": [
    "nls_mean_result = knn_cv_wrapper(\n",
    "    labels=df['neal_label'].values, years=df['year'].values, nls=df['nl'].values, folds=ooc_folds,\n",
    "    savepath=None)\n",
    "test_preds = nls_mean_result[1]\n",
    "for country in COUNTRIES:\n",
    "    indices = country_indices[country]\n",
    "    evaluate(labels=df.loc[indices, 'neal_label'], preds=test_preds[indices], do_print=True, title=country)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fold A, chosen k: 50\n",
      "fold B, chosen k: 50\n",
      "fold C, chosen k: 200\n",
      "fold D, chosen k: 30\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_knn_val_performance(nls_mean_result, labels, title='NL mean')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7fab0cb4ee80>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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.scatter(df['neal_label'], test_preds, c=df['country'].map(COUNTRIES.index), cmap='tab10')"
   ]
  }
 ],
 "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"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": true,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}