j40-cejst-2/data/data-pipeline/data_pipeline/ipython/agricultural_loss_indicator.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "dc8a46ce-3dbf-49ee-a0ab-1449fd6d176d",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import sys\n",
    "from pathlib import Path\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "sys.path.append(\"../../data_pipeline/\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ebdcdf20-08b6-48e6-b28b-4bebbb3655c2",
   "metadata": {},
   "source": [
    "# Examining agricultural loss indicator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "78cf4d99-6096-43a8-95e2-4e8328a78b18",
   "metadata": {},
   "outputs": [],
   "source": [
    "DATA_PATH = Path.cwd().parent / \"data\"\n",
    "\n",
    "urban_rural_from_geocorr = pd.read_csv(\n",
    "    DATA_PATH / \"dataset/geocorr/usa.csv\",\n",
    "    dtype={\"GEOID10_TRACT\": str},\n",
    ")\n",
    "\n",
    "score_m = pd.read_csv(\n",
    "    DATA_PATH / \"score/csv/full/usa.csv\",\n",
    "    dtype={\"GEOID10_TRACT\": str},\n",
    "    usecols=[\n",
    "        \"Expected agricultural loss rate (Natural Hazards Risk Index)\",\n",
    "        \"Expected agricultural loss rate (Natural Hazards Risk Index) (percentile)\",\n",
    "        \"Greater than or equal to the 90th percentile for expected agriculture loss rate, is low income, and has a low percent of higher ed students?\",\n",
    "        \"Urban Heuristic Flag\",\n",
    "        \"Is low income and has a low percent of higher ed students?\",\n",
    "        \"GEOID10_TRACT\",\n",
    "        \"Total threshold criteria exceeded\",\n",
    "    ],\n",
    ")\n",
    "\n",
    "# Note that I downloaded this fresh because I am paranoid; this is not going to load on other computers and I am sorry!\n",
    "nri_full = pd.read_csv(\n",
    "    \"/Users/emmausds/Desktop/current-work/NRI_Table_CensusTracts.csv\",\n",
    "    dtype={\"TRACTFIPS\": str},\n",
    "    usecols=[\n",
    "        \"TRACTFIPS\",\n",
    "        \"AGRIVALUE\",\n",
    "        \"CWAV_EALA\",\n",
    "        \"DRGT_EALA\",\n",
    "        \"HAIL_EALA\",\n",
    "        \"HWAV_EALA\",\n",
    "        \"HRCN_EALA\",\n",
    "        \"RFLD_EALA\",\n",
    "        \"SWND_EALA\",\n",
    "        \"TRND_EALA\",\n",
    "        \"WFIR_EALA\",\n",
    "        \"WNTW_EALA\",\n",
    "    ],\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "04ed5585-e7e5-4209-8e55-75a3e1fce633",
   "metadata": {},
   "source": [
    "## Understanding our current implementation\n",
    "\n",
    "In our current implementation, on average, urban areas have a higher NRI (scaled) and a higher percentile of the loss rate. The share of Rural and Urban tracts identified by this threshold is roughly equal. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "48250479-4074-4b04-9f8b-56a6105e8225",
   "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>Expected agricultural loss rate (Natural Hazards Risk Index)</th>\n",
       "      <th>Expected agricultural loss rate (Natural Hazards Risk Index) (percentile)</th>\n",
       "      <th>Greater than or equal to the 90th percentile for expected agriculture loss rate, is low income, and has a low percent of higher ed students?</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Urban Heuristic Flag</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.0</th>\n",
       "      <td>0.011502</td>\n",
       "      <td>0.486045</td>\n",
       "      <td>0.021132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0</th>\n",
       "      <td>0.016255</td>\n",
       "      <td>0.505159</td>\n",
       "      <td>0.018296</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      Expected agricultural loss rate (Natural Hazards Risk Index)  \\\n",
       "Urban Heuristic Flag                                                                 \n",
       "0.0                                                            0.011502              \n",
       "1.0                                                            0.016255              \n",
       "\n",
       "                      Expected agricultural loss rate (Natural Hazards Risk Index) (percentile)  \\\n",
       "Urban Heuristic Flag                                                                              \n",
       "0.0                                                            0.486045                           \n",
       "1.0                                                            0.505159                           \n",
       "\n",
       "                      Greater than or equal to the 90th percentile for expected agriculture loss rate, is low income, and has a low percent of higher ed students?  \n",
       "Urban Heuristic Flag                                                                                                                                                \n",
       "0.0                                                            0.021132                                                                                             \n",
       "1.0                                                            0.018296                                                                                             "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score_m.groupby(\"Urban Heuristic Flag\")[\n",
    "    [\n",
    "        \"Expected agricultural loss rate (Natural Hazards Risk Index)\",\n",
    "        \"Expected agricultural loss rate (Natural Hazards Risk Index) (percentile)\",\n",
    "        \"Greater than or equal to the 90th percentile for expected agriculture loss rate, is low income, and has a low percent of higher ed students?\",\n",
    "    ]\n",
    "].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "da540db6-d07b-4dac-9959-23e29df0881b",
   "metadata": {},
   "source": [
    "We can also look at the distribution of percentiles among the urban and rural tracts. This is very much not what I might expect -- I'd hope the rural areas were \"flatter\" in distribution than the urban areas.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7adf50e6-9293-40df-afd2-b1fe152d88ad",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.9/site-packages/seaborn/distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n",
      "/usr/local/lib/python3.9/site-packages/seaborn/distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "tmp = sns.FacetGrid(\n",
    "    data=score_m, col=\"Urban Heuristic Flag\", col_wrap=2, height=7\n",
    ")\n",
    "tmp.map(\n",
    "    sns.distplot,\n",
    "    \"Expected agricultural loss rate (Natural Hazards Risk Index) (percentile)\",\n",
    "    bins=20,\n",
    "    kde=True,\n",
    "    color=\"#62acef\",\n",
    ")\n",
    "tmp.set(xlim=(0, 1.0))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1bff0654-6a88-4511-b17f-492d41023d9f",
   "metadata": {},
   "source": [
    "But, if we look at just the raw loss rates, we see a very different distribution. This suggests to me that we are perhaps elevating the urban areas in our wealth-neutral metric. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "b10adc64-3dae-429b-a0d3-b54f701587df",
   "metadata": {},
   "outputs": [],
   "source": [
    "nri_with_flag = (\n",
    "    nri_full.set_index(\"TRACTFIPS\")\n",
    "    .merge(\n",
    "        urban_rural_from_geocorr.set_index(\"GEOID10_TRACT\"),\n",
    "        left_index=True,\n",
    "        right_index=True,\n",
    "        how=\"left\",\n",
    "    )\n",
    "    .reset_index()\n",
    ")\n",
    "\n",
    "nri_with_flag[\"total_ag_loss\"] = nri_with_flag.filter(like=\"EALA\").sum(axis=1)\n",
    "nri_with_flag[\"total_ag_loss_pctile\"] = nri_with_flag[\"total_ag_loss\"].rank(\n",
    "    pct=True\n",
    ")\n",
    "\n",
    "nri_with_flag.groupby(\"Urban Heuristic Flag\")[\"total_ag_loss_pctile\"].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae3bd6f6-431a-4b73-a81d-a3e4eb14d708",
   "metadata": {},
   "source": [
    "When we look at the distribution of agricultural value in census tracts that are urban and rural, we see that the agricultural value for urban tracts is quite skewed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "38a8e0a0-0cc4-43c9-b0bb-2062da59a639",
   "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>0.00</th>\n",
       "      <th>0.10</th>\n",
       "      <th>0.20</th>\n",
       "      <th>0.30</th>\n",
       "      <th>0.40</th>\n",
       "      <th>0.50</th>\n",
       "      <th>0.60</th>\n",
       "      <th>0.70</th>\n",
       "      <th>0.80</th>\n",
       "      <th>0.90</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Urban Heuristic Flag</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.00</th>\n",
       "      <td>0.00</td>\n",
       "      <td>407,226.75</td>\n",
       "      <td>1,305,567.65</td>\n",
       "      <td>2,586,000.00</td>\n",
       "      <td>4,411,909.10</td>\n",
       "      <td>7,385,163.62</td>\n",
       "      <td>11,582,526.63</td>\n",
       "      <td>18,205,814.38</td>\n",
       "      <td>30,706,464.84</td>\n",
       "      <td>56,067,604.96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.00</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>141.35</td>\n",
       "      <td>1,812.91</td>\n",
       "      <td>9,755.31</td>\n",
       "      <td>44,319.94</td>\n",
       "      <td>202,972.50</td>\n",
       "      <td>1,060,793.40</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      0.00       0.10         0.20         0.30         0.40  \\\n",
       "Urban Heuristic Flag                                                           \n",
       "0.00                  0.00 407,226.75 1,305,567.65 2,586,000.00 4,411,909.10   \n",
       "1.00                  0.00       0.00         0.00         0.00       141.35   \n",
       "\n",
       "                             0.50          0.60          0.70          0.80  \\\n",
       "Urban Heuristic Flag                                                          \n",
       "0.00                 7,385,163.62 11,582,526.63 18,205,814.38 30,706,464.84   \n",
       "1.00                     1,812.91      9,755.31     44,319.94    202,972.50   \n",
       "\n",
       "                              0.90  \n",
       "Urban Heuristic Flag                \n",
       "0.00                 56,067,604.96  \n",
       "1.00                  1,060,793.40  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.options.display.float_format = \"{:,.2f}\".format\n",
    "nri_with_flag.groupby(\"Urban Heuristic Flag\")[\"AGRIVALUE\"].quantile(\n",
    "    q=np.arange(0, 1, step=0.1)\n",
    ").unstack()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8aed6062-9503-4120-bab8-f840cb524365",
   "metadata": {},
   "source": [
    "## Updating the metric\n",
    "\n",
    "So we clip the values such that agrivalue is defined as the maximum of the tract's agricultural value AND the 10th percentile of rural tracts' agrivalues. When we do that, we see that a lot more (proportionally) rural tracts exceed the 90th percentile. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "2b3a3b3e-76c2-4b6b-b96b-d439580de592",
   "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>Urban Heuristic Flag</th>\n",
       "      <th>0.00</th>\n",
       "      <th>1.00</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Expected agricultural loss rate (Natural Hazards Risk Index) exceeds 90th percentile, adjusted</th>\n",
       "      <td>0.24</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Expected agricultural loss rate (Natural Hazards Risk Index) (percentile, adjusted)</th>\n",
       "      <td>0.77</td>\n",
       "      <td>0.43</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Urban Heuristic Flag                                0.00  1.00\n",
       "Expected agricultural loss rate (Natural Hazard...  0.24  0.07\n",
       "Expected agricultural loss rate (Natural Hazard...  0.77  0.43"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_left_clip = nri_with_flag[nri_with_flag[\"Urban Heuristic Flag\"] == 0][\n",
    "    \"AGRIVALUE\"\n",
    "].quantile(0.1)\n",
    "\n",
    "nri_with_flag[\n",
    "    \"Expected agricultural loss rate (Natural Hazards Risk Index) (percentile, adjusted)\"\n",
    "] = (\n",
    "    nri_with_flag[\"total_ag_loss\"]\n",
    "    / (nri_with_flag[\"AGRIVALUE\"].clip(lower=final_left_clip))\n",
    ").rank(\n",
    "    pct=True\n",
    ")\n",
    "\n",
    "nri_with_flag[\n",
    "    \"Expected agricultural loss rate (Natural Hazards Risk Index) exceeds 90th percentile, adjusted\"\n",
    "] = (\n",
    "    nri_with_flag[\n",
    "        \"Expected agricultural loss rate (Natural Hazards Risk Index) (percentile, adjusted)\"\n",
    "    ]\n",
    "    >= 0.9\n",
    ")\n",
    "\n",
    "nri_with_flag.groupby(\"Urban Heuristic Flag\")[\n",
    "    [\n",
    "        \"Expected agricultural loss rate (Natural Hazards Risk Index) exceeds 90th percentile, adjusted\",\n",
    "        \"Expected agricultural loss rate (Natural Hazards Risk Index) (percentile, adjusted)\",\n",
    "    ]\n",
    "].mean().T"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "54585b02-d26e-4986-8a8a-9828582e7ac1",
   "metadata": {},
   "source": [
    "The percentile distribution look a little better to me (but like, not fantastic)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "4fd90ace-65f7-4fa1-973f-922b56939ecd",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.9/site-packages/seaborn/distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n",
      "/usr/local/lib/python3.9/site-packages/seaborn/distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "tmp = sns.FacetGrid(\n",
    "    data=nri_with_flag, col=\"Urban Heuristic Flag\", col_wrap=1, height=4\n",
    ")\n",
    "tmp.map(\n",
    "    sns.distplot,\n",
    "    \"Expected agricultural loss rate (Natural Hazards Risk Index) (percentile, adjusted)\",\n",
    "    bins=20,\n",
    "    kde=True,\n",
    "    color=\"#62acef\",\n",
    ")\n",
    "tmp.set(xlim=(0, 1.0))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a273e5b1-c4a0-45b3-8f57-3be116685c10",
   "metadata": {},
   "source": [
    "We can also look at the distribution of actually exceeding the threshold -- so also adding in the socioeconomic information from Score M. Note that *some tracts* are not shared between score M and the NRI data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "36670c86-230a-491b-bf44-767c1e1cd779",
   "metadata": {},
   "outputs": [],
   "source": [
    "nri_with_flag[\"low_inc_low_highed\"] = (\n",
    "    nri_with_flag[\"TRACTFIPS\"]\n",
    "    .map(\n",
    "        score_m.set_index(\"GEOID10_TRACT\")[\n",
    "            \"Is low income and has a low percent of higher ed students?\"\n",
    "        ].to_dict()\n",
    "    )\n",
    "    .fillna(False)\n",
    ")\n",
    "\n",
    "nri_with_flag[\"would_exceed_threshold\"] = (\n",
    "    nri_with_flag[\"low_inc_low_highed\"]\n",
    "    & nri_with_flag[\n",
    "        \"Expected agricultural loss rate (Natural Hazards Risk Index) exceeds 90th percentile, adjusted\"\n",
    "    ]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "bdb9a4b3-954a-4389-86d0-84d896c1679c",
   "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>sum</th>\n",
       "      <th>mean</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Urban Heuristic Flag</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.00000</th>\n",
       "      <td>1190</td>\n",
       "      <td>0.08439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.00000</th>\n",
       "      <td>1158</td>\n",
       "      <td>0.01982</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       sum    mean\n",
       "Urban Heuristic Flag              \n",
       "0.00000               1190 0.08439\n",
       "1.00000               1158 0.01982"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.options.display.float_format = \"{:.5f}\".format\n",
    "nri_with_flag.groupby(\"Urban Heuristic Flag\")[\"would_exceed_threshold\"].agg(\n",
    "    [\"sum\", \"mean\"]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "015de431-0c39-49b7-9144-1bfa776e4f75",
   "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>sum</th>\n",
       "      <th>mean</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Urban Heuristic Flag</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.00000</th>\n",
       "      <td>298</td>\n",
       "      <td>0.02113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.00000</th>\n",
       "      <td>1069</td>\n",
       "      <td>0.01830</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       sum    mean\n",
       "Urban Heuristic Flag              \n",
       "0.00000                298 0.02113\n",
       "1.00000               1069 0.01830"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score_m.groupby(\"Urban Heuristic Flag\")[\n",
    "    \"Greater than or equal to the 90th percentile for expected agriculture loss rate, is low income, and has a low percent of higher ed students?\"\n",
    "].agg([\"sum\", \"mean\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fd175407-07ee-4712-ac7e-ddd06ca8c622",
   "metadata": {},
   "source": [
    "This really increases the number of tracts identified - from just under 2% to just over 3% - and adds 259 tracts, removes 116."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "9c164dfc-9053-462b-8a55-1b802694ec48",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2348"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "1367"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "259"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(\n",
    "    nri_with_flag[\"would_exceed_threshold\"].sum(),\n",
    "    score_m[\n",
    "        \"Greater than or equal to the 90th percentile for expected agriculture loss rate, is low income, and has a low percent of higher ed students?\"\n",
    "    ].sum(),\n",
    ")\n",
    "\n",
    "all_ag_loss_tracts = nri_with_flag[nri_with_flag[\"would_exceed_threshold\"]][\n",
    "    \"TRACTFIPS\"\n",
    "].unique()\n",
    "\n",
    "all_scorem_tracts = score_m[score_m[\"Total threshold criteria exceeded\"] > 0][\n",
    "    \"GEOID10_TRACT\"\n",
    "].unique()\n",
    "\n",
    "display(len(set(all_ag_loss_tracts).difference(all_scorem_tracts)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "9a225512-758c-4807-a598-bc8c97e20a47",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "116"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "score_m[\"adjusted\"] = score_m[\"Total threshold criteria exceeded\"] - score_m[\n",
    "    \"Greater than or equal to the 90th percentile for expected agriculture loss rate, is low income, and has a low percent of higher ed students?\"\n",
    "].astype(int)\n",
    "\n",
    "score_m_adjusted_tracts = set(\n",
    "    score_m[score_m[\"adjusted\"] > 0][\"GEOID10_TRACT\"]\n",
    ").union(all_ag_loss_tracts)\n",
    "display(len(set(all_scorem_tracts).difference(score_m_adjusted_tracts)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "73b11c1d-5661-48e3-88f3-255b0ef21768",
   "metadata": {},
   "source": [
    "## Scratch \n",
    "Choosing a left clip value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "01e50420-a82c-42ee-bf2b-a90fdb61bcef",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "At threshold 0.00, minimum value is $0\n",
      "At threshold 0.05, minimum value is $82,341\n",
      "At threshold 0.10, minimum value is $407,227\n",
      "At threshold 0.15, minimum value is $822,946\n",
      "At threshold 0.20, minimum value is $1,305,568\n",
      "At threshold 0.25, minimum value is $1,881,536\n",
      "At threshold 0.30, minimum value is $2,586,000\n",
      "At threshold 0.35, minimum value is $3,375,394\n",
      "At threshold 0.40, minimum value is $4,411,909\n",
      "At threshold 0.45, minimum value is $5,719,970\n",
      "At threshold 0.50, minimum value is $7,385,164\n",
      "At threshold 0.55, minimum value is $9,228,310\n",
      "At threshold 0.60, minimum value is $11,582,527\n",
      "At threshold 0.65, minimum value is $14,421,393\n",
      "At threshold 0.70, minimum value is $18,205,814\n",
      "At threshold 0.75, minimum value is $23,497,859\n",
      "At threshold 0.80, minimum value is $30,706,465\n",
      "At threshold 0.85, minimum value is $40,827,282\n",
      "At threshold 0.90, minimum value is $56,067,605\n",
      "At threshold 0.95, minimum value is $86,129,868\n"
     ]
    }
   ],
   "source": [
    "for threshold in np.arange(0, 1, 0.05):\n",
    "    left_clip = nri_with_flag[nri_with_flag[\"Urban Heuristic Flag\"] == 0][\n",
    "        \"AGRIVALUE\"\n",
    "    ].quantile(threshold)\n",
    "    print(\n",
    "        \"At threshold {:.2f}, minimum value is ${:,.0f}\".format(\n",
    "            threshold, left_clip\n",
    "        )\n",
    "    )\n",
    "    tmp_value = nri_with_flag[\"AGRIVALUE\"].clip(lower=left_clip)\n",
    "    nri_with_flag[\"total_ag_loss_pctile_{:.2f}\".format(threshold)] = (\n",
    "        nri_with_flag[\"total_ag_loss\"] / tmp_value\n",
    "    ).rank(pct=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "0bcddf71-4224-4e7d-b6f7-1edaeba5893a",
   "metadata": {},
   "outputs": [],
   "source": [
    "tmp = (\n",
    "    nri_with_flag.groupby(\"Urban Heuristic Flag\")[\n",
    "        [\n",
    "            \"total_ag_loss_pctile\",\n",
    "            \"total_ag_loss_pctile_0.00\",\n",
    "            \"total_ag_loss_pctile_0.05\",\n",
    "            \"total_ag_loss_pctile_0.10\",\n",
    "            \"total_ag_loss_pctile_0.15\",\n",
    "            \"total_ag_loss_pctile_0.20\",\n",
    "            \"total_ag_loss_pctile_0.25\",\n",
    "            \"total_ag_loss_pctile_0.30\",\n",
    "            \"total_ag_loss_pctile_0.35\",\n",
    "            \"total_ag_loss_pctile_0.40\",\n",
    "            \"total_ag_loss_pctile_0.45\",\n",
    "            \"total_ag_loss_pctile_0.50\",\n",
    "            \"total_ag_loss_pctile_0.55\",\n",
    "            \"total_ag_loss_pctile_0.60\",\n",
    "            \"total_ag_loss_pctile_0.65\",\n",
    "            \"total_ag_loss_pctile_0.70\",\n",
    "            \"total_ag_loss_pctile_0.75\",\n",
    "            \"total_ag_loss_pctile_0.80\",\n",
    "            \"total_ag_loss_pctile_0.85\",\n",
    "            \"total_ag_loss_pctile_0.90\",\n",
    "            \"total_ag_loss_pctile_0.95\",\n",
    "        ]\n",
    "    ]\n",
    "    .mean()\n",
    "    .T\n",
    ")\n",
    "\n",
    "left_pctile = tmp.reset_index()[\"index\"].str.split(\"_\", expand=True)[4]\n",
    "graph_data = (\n",
    "    pd.concat([tmp.reset_index(), left_pctile], axis=1)\n",
    "    .rename(\n",
    "        columns={\n",
    "            0: \"Rural\",\n",
    "            1: \"Urban\",\n",
    "            4: \"Left clip value\",\n",
    "        }\n",
    "    )\n",
    "    .set_index(\"Left clip value\")[[\"Rural\", \"Urban\"]]\n",
    "    .stack()\n",
    "    .reset_index()\n",
    "    .rename(\n",
    "        columns={\"level_1\": \"Tract classification\", 0: \"Average percentile\"}\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "12fb9323-e932-4392-b722-8fca3c127b0e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Left clip value', ylabel='Average percentile'>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(11, 7))\n",
    "sns.lineplot(\n",
    "    x=\"Left clip value\",\n",
    "    y=\"Average percentile\",\n",
    "    hue=\"Tract classification\",\n",
    "    data=graph_data,\n",
    "    palette=\"colorblind\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c898c542-1dad-43b7-b20b-ee065f96c227",
   "metadata": {},
   "source": [
    "Note -- some tracts have a missing urban / rural flag. They get left out from the means. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "adfcdbe0-fdcc-4ec7-8788-5628641bed8b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0    58429\n",
       "0.0    14101\n",
       "NaN      209\n",
       "Name: Urban Heuristic Flag, dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nri_with_flag[\"Urban Heuristic Flag\"].value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0fba097c-b5e0-4511-b072-05511906dd9b",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}