pull_show_feature_imps.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import plotly\n",
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "orig = pd.read_csv('./output_files/OutcomeReport_ttt_FeatureImportInfoRBGinsburg.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>ID</th>\n",
       "      <th>decisionType_1</th>\n",
       "      <th>decisionType_2</th>\n",
       "      <th>decisionType_4</th>\n",
       "      <th>decisionType_6</th>\n",
       "      <th>decisionType_7</th>\n",
       "      <th>naturalCourt_1301</th>\n",
       "      <th>naturalCourt_1303</th>\n",
       "      <th>naturalCourt_1401</th>\n",
       "      <th>...</th>\n",
       "      <th>issue_140030.0</th>\n",
       "      <th>lawType_1.0</th>\n",
       "      <th>lawType_2.0</th>\n",
       "      <th>lawType_3.0</th>\n",
       "      <th>lawType_4.0</th>\n",
       "      <th>lawType_5.0</th>\n",
       "      <th>lawType_6.0</th>\n",
       "      <th>lawType_8.0</th>\n",
       "      <th>lawType_9.0</th>\n",
       "      <th>lawType_999.0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>RBGinsburg-RFT-31</td>\n",
       "      <td>0.022286</td>\n",
       "      <td>0.029797</td>\n",
       "      <td>0.000373</td>\n",
       "      <td>0.016017</td>\n",
       "      <td>0.011715</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000361</td>\n",
       "      <td>0.004025</td>\n",
       "      <td>0.017549</td>\n",
       "      <td>0.009433</td>\n",
       "      <td>0.014453</td>\n",
       "      <td>0.002717</td>\n",
       "      <td>0.010522</td>\n",
       "      <td>0.000106</td>\n",
       "      <td>0.003263</td>\n",
       "      <td>0.007286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>RBGinsburg-ADA-31</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.136667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.036667</td>\n",
       "      <td>0.073333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.106667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.030000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 719 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0                 ID  decisionType_1  decisionType_2  \\\n",
       "0           0  RBGinsburg-RFT-31        0.022286        0.029797   \n",
       "1           1  RBGinsburg-ADA-31        0.000000        0.136667   \n",
       "\n",
       "   decisionType_4  decisionType_6  decisionType_7  naturalCourt_1301  \\\n",
       "0        0.000373        0.016017        0.011715                0.0   \n",
       "1        0.000000        0.036667        0.073333                0.0   \n",
       "\n",
       "   naturalCourt_1303  naturalCourt_1401      ...        issue_140030.0  \\\n",
       "0                0.0                0.0      ...              0.000361   \n",
       "1                0.0                0.0      ...              0.000000   \n",
       "\n",
       "   lawType_1.0  lawType_2.0  lawType_3.0  lawType_4.0  lawType_5.0  \\\n",
       "0     0.004025     0.017549     0.009433     0.014453     0.002717   \n",
       "1     0.000000     0.106667     0.000000     0.030000     0.000000   \n",
       "\n",
       "   lawType_6.0  lawType_8.0  lawType_9.0  lawType_999.0  \n",
       "0     0.010522     0.000106     0.003263       0.007286  \n",
       "1     0.000000     0.000000     0.000000       0.000000  \n",
       "\n",
       "[2 rows x 719 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "orig.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = orig.drop(['Unnamed: 0', 'ID'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>decisionType_1</th>\n",
       "      <th>decisionType_2</th>\n",
       "      <th>decisionType_4</th>\n",
       "      <th>decisionType_6</th>\n",
       "      <th>decisionType_7</th>\n",
       "      <th>naturalCourt_1301</th>\n",
       "      <th>naturalCourt_1303</th>\n",
       "      <th>naturalCourt_1401</th>\n",
       "      <th>naturalCourt_1402</th>\n",
       "      <th>naturalCourt_1403</th>\n",
       "      <th>...</th>\n",
       "      <th>issue_140030.0</th>\n",
       "      <th>lawType_1.0</th>\n",
       "      <th>lawType_2.0</th>\n",
       "      <th>lawType_3.0</th>\n",
       "      <th>lawType_4.0</th>\n",
       "      <th>lawType_5.0</th>\n",
       "      <th>lawType_6.0</th>\n",
       "      <th>lawType_8.0</th>\n",
       "      <th>lawType_9.0</th>\n",
       "      <th>lawType_999.0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.022286</td>\n",
       "      <td>0.029797</td>\n",
       "      <td>0.000373</td>\n",
       "      <td>0.016017</td>\n",
       "      <td>0.011715</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000361</td>\n",
       "      <td>0.004025</td>\n",
       "      <td>0.017549</td>\n",
       "      <td>0.009433</td>\n",
       "      <td>0.014453</td>\n",
       "      <td>0.002717</td>\n",
       "      <td>0.010522</td>\n",
       "      <td>0.000106</td>\n",
       "      <td>0.003263</td>\n",
       "      <td>0.007286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.136667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.036667</td>\n",
       "      <td>0.073333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.106667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.030000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 717 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   decisionType_1  decisionType_2  decisionType_4  decisionType_6  \\\n",
       "0        0.022286        0.029797        0.000373        0.016017   \n",
       "1        0.000000        0.136667        0.000000        0.036667   \n",
       "\n",
       "   decisionType_7  naturalCourt_1301  naturalCourt_1303  naturalCourt_1401  \\\n",
       "0        0.011715                0.0                0.0                0.0   \n",
       "1        0.073333                0.0                0.0                0.0   \n",
       "\n",
       "   naturalCourt_1402  naturalCourt_1403      ...        issue_140030.0  \\\n",
       "0                0.0                0.0      ...              0.000361   \n",
       "1                0.0                0.0      ...              0.000000   \n",
       "\n",
       "   lawType_1.0  lawType_2.0  lawType_3.0  lawType_4.0  lawType_5.0  \\\n",
       "0     0.004025     0.017549     0.009433     0.014453     0.002717   \n",
       "1     0.000000     0.106667     0.000000     0.030000     0.000000   \n",
       "\n",
       "   lawType_6.0  lawType_8.0  lawType_9.0  lawType_999.0  \n",
       "0     0.010522     0.000106     0.003263       0.007286  \n",
       "1     0.000000     0.000000     0.000000       0.000000  \n",
       "\n",
       "[2 rows x 717 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.drop([1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>decisionType_1</th>\n",
       "      <th>decisionType_2</th>\n",
       "      <th>decisionType_4</th>\n",
       "      <th>decisionType_6</th>\n",
       "      <th>decisionType_7</th>\n",
       "      <th>naturalCourt_1301</th>\n",
       "      <th>naturalCourt_1303</th>\n",
       "      <th>naturalCourt_1401</th>\n",
       "      <th>naturalCourt_1402</th>\n",
       "      <th>naturalCourt_1403</th>\n",
       "      <th>...</th>\n",
       "      <th>issue_140030.0</th>\n",
       "      <th>lawType_1.0</th>\n",
       "      <th>lawType_2.0</th>\n",
       "      <th>lawType_3.0</th>\n",
       "      <th>lawType_4.0</th>\n",
       "      <th>lawType_5.0</th>\n",
       "      <th>lawType_6.0</th>\n",
       "      <th>lawType_8.0</th>\n",
       "      <th>lawType_9.0</th>\n",
       "      <th>lawType_999.0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.022286</td>\n",
       "      <td>0.029797</td>\n",
       "      <td>0.000373</td>\n",
       "      <td>0.016017</td>\n",
       "      <td>0.011715</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000361</td>\n",
       "      <td>0.004025</td>\n",
       "      <td>0.017549</td>\n",
       "      <td>0.009433</td>\n",
       "      <td>0.014453</td>\n",
       "      <td>0.002717</td>\n",
       "      <td>0.010522</td>\n",
       "      <td>0.000106</td>\n",
       "      <td>0.003263</td>\n",
       "      <td>0.007286</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 717 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   decisionType_1  decisionType_2  decisionType_4  decisionType_6  \\\n",
       "0        0.022286        0.029797        0.000373        0.016017   \n",
       "\n",
       "   decisionType_7  naturalCourt_1301  naturalCourt_1303  naturalCourt_1401  \\\n",
       "0        0.011715                0.0                0.0                0.0   \n",
       "\n",
       "   naturalCourt_1402  naturalCourt_1403      ...        issue_140030.0  \\\n",
       "0                0.0                0.0      ...              0.000361   \n",
       "\n",
       "   lawType_1.0  lawType_2.0  lawType_3.0  lawType_4.0  lawType_5.0  \\\n",
       "0     0.004025     0.017549     0.009433     0.014453     0.002717   \n",
       "\n",
       "   lawType_6.0  lawType_8.0  lawType_9.0  lawType_999.0  \n",
       "0     0.010522     0.000106     0.003263       0.007286  \n",
       "\n",
       "[1 rows x 717 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.transpose()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.columns = ['feature_importance']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.sort_values(by = ['feature_importance'], ascending = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>feature_importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>lcDispositionDirection_2.0</th>\n",
       "      <td>0.047282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lcDispositionDirection_1.0</th>\n",
       "      <td>0.044473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>decisionType_2</th>\n",
       "      <td>0.029797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>issue_90150.0</th>\n",
       "      <td>0.022549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>decisionType_1</th>\n",
       "      <td>0.022286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lawType_2.0</th>\n",
       "      <td>0.017549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>issue_20360.0</th>\n",
       "      <td>0.016091</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>decisionType_6</th>\n",
       "      <td>0.016017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lawType_4.0</th>\n",
       "      <td>0.014453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lcDisposition_2.0</th>\n",
       "      <td>0.013430</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>certReason_12.0</th>\n",
       "      <td>0.013168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>decisionType_7</th>\n",
       "      <td>0.011715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lcDisagreement_0.0</th>\n",
       "      <td>0.010964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lawType_6.0</th>\n",
       "      <td>0.010522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lcDisagreement_1.0</th>\n",
       "      <td>0.010503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>caseSource_999.0</th>\n",
       "      <td>0.010197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>chief_Rehnquist</th>\n",
       "      <td>0.010126</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>certReason_2.0</th>\n",
       "      <td>0.009695</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>naturalCourt_1607</th>\n",
       "      <td>0.009693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>chief_Roberts</th>\n",
       "      <td>0.009508</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            feature_importance\n",
       "lcDispositionDirection_2.0            0.047282\n",
       "lcDispositionDirection_1.0            0.044473\n",
       "decisionType_2                        0.029797\n",
       "issue_90150.0                         0.022549\n",
       "decisionType_1                        0.022286\n",
       "lawType_2.0                           0.017549\n",
       "issue_20360.0                         0.016091\n",
       "decisionType_6                        0.016017\n",
       "lawType_4.0                           0.014453\n",
       "lcDisposition_2.0                     0.013430\n",
       "certReason_12.0                       0.013168\n",
       "decisionType_7                        0.011715\n",
       "lcDisagreement_0.0                    0.010964\n",
       "lawType_6.0                           0.010522\n",
       "lcDisagreement_1.0                    0.010503\n",
       "caseSource_999.0                      0.010197\n",
       "chief_Rehnquist                       0.010126\n",
       "certReason_2.0                        0.009695\n",
       "naturalCourt_1607                     0.009693\n",
       "chief_Roberts                         0.009508"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "N = 20\n",
    "features = df.index.values[0:N]\n",
    "y_pos = np.arange(N)\n",
    "imps = df.feature_importance.values[0:N]\n",
    "\n",
    "plt.figure(figsize=(15,8))\n",
    "plt.bar(y_pos, imps, align='center', alpha=0.5)\n",
    "plt.xticks(y_pos, features, rotation = 90)\n",
    "h = plt.ylabel('Feature Codebook Reference')\n",
    "plt.title('Feature Importances for RBGinsburg')\n",
    "\n",
    "plt.draw()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "717"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(imps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "19"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}