Introspect Trees (Depth).ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 1\n",
    "import pandas as pd\n",
    "import os\n",
    "import sys\n",
    "scriptpath = \"binarygridsearch.py\"\n",
    "# Do the import\n",
    "sys.path.append(os.path.abspath(scriptpath))\n",
    "# Do the import\n",
    "import binarygridsearch as bgs\n",
    "%aimport binarygridsearch\n",
    "\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "from sklearn.datasets import load_breast_cancer\n",
    "from sklearn.tree import _tree\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#From here: https://aysent.github.io/2015/11/08/random-forest-leaf-visualization.html\n",
    "def leaf_depths(tree, node_id = 0):\n",
    "    \n",
    "    '''\n",
    "    tree.children_left and tree.children_right store ids\n",
    "    of left and right chidren of a given node\n",
    "    '''\n",
    "    left_child = tree.children_left[node_id]\n",
    "    right_child = tree.children_right[node_id]\n",
    "\n",
    "    '''\n",
    "    If a given node is terminal, \n",
    "    both left and right children are set to _tree.TREE_LEAF\n",
    "    '''\n",
    "    if left_child == _tree.TREE_LEAF:\n",
    "        \n",
    "        '''\n",
    "        Set depth of terminal nodes to 0\n",
    "        '''\n",
    "        depths = np.array([0])\n",
    "    else:\n",
    "        '''\n",
    "        Get depths of left and right children and\n",
    "        increment them by 1\n",
    "        '''\n",
    "        left_depths = leaf_depths(tree, left_child) + 1\n",
    "        right_depths = leaf_depths(tree, right_child) + 1\n",
    " \n",
    "        depths = np.append(left_depths, right_depths)\n",
    " \n",
    "    return depths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getForestDepths(X, y, kwargs):\n",
    "    clf = RandomForestClassifier(**kwargs)\n",
    "    clf.fit(X, y)\n",
    "    allDepths = [leaf_depths(estimator.tree_) \n",
    "                 for estimator in clf.estimators_]\n",
    "    return {\"min\": np.hstack(allDepths).min(),\n",
    "           \"max\": np.hstack(allDepths).max()}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = load_breast_cancer()\n",
    "X, y = data.data, data.target\n",
    "\n",
    "rfArgs = {\"random_state\": 0,\n",
    "          \"n_jobs\": -1,\n",
    "          \"class_weight\": \"balanced\",\n",
    "         \"n_estimators\": 18,\n",
    "         \"oob_score\": True}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'min': 2, 'max': 9}"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "getForestDepths(X, y, rfArgs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getForestAccuracy(X, y, metric, kwargs):\n",
    "    clf = RandomForestClassifier(**kwargs)\n",
    "    clf.fit(X, y)\n",
    "    y_pred = clf.oob_decision_function_[:, 1]\n",
    "    return metric(y, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>max_depth</th>\n",
       "      <th>score</th>\n",
       "      <th>time</th>\n",
       "      <th>scoreTimeRatio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>0.980531</td>\n",
       "      <td>1.037448</td>\n",
       "      <td>0.009858</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9</td>\n",
       "      <td>0.750646</td>\n",
       "      <td>0.154658</td>\n",
       "      <td>0.138930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>1.158169</td>\n",
       "      <td>0.037448</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0.158169</td>\n",
       "      <td>0.244608</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   max_depth     score      time  scoreTimeRatio\n",
       "0          2  0.980531  1.037448        0.009858\n",
       "1          9  0.750646  0.154658        0.138930\n",
       "2          6  1.158169  0.037448        1.000000\n",
       "3          4  0.158169  0.244608        0.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>max_depth</th>\n",
       "      <th>score</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>0.982189</td>\n",
       "      <td>0.155940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9</td>\n",
       "      <td>0.980736</td>\n",
       "      <td>0.132366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>0.983312</td>\n",
       "      <td>0.129236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0.976990</td>\n",
       "      <td>0.134768</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   max_depth     score      time\n",
       "0          2  0.982189  0.155940\n",
       "1          9  0.980736  0.132366\n",
       "2          6  0.983312  0.129236\n",
       "3          4  0.976990  0.134768"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.vegalite.v2+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v2.4.3.json",
       "config": {
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "values": [
         {
          "max_depth": 2,
          "value": 0.9805308254963617,
          "variable": "score"
         },
         {
          "max_depth": 9,
          "value": 0.7506457680250784,
          "variable": "score"
         },
         {
          "max_depth": 6,
          "value": 1.1581692789968638,
          "variable": "score"
         },
         {
          "max_depth": 4,
          "value": 0.15816927899686378,
          "variable": "score"
         },
         {
          "max_depth": 2,
          "value": 1.03744757339741,
          "variable": "time"
         },
         {
          "max_depth": 9,
          "value": 0.15465847809932678,
          "variable": "time"
         },
         {
          "max_depth": 6,
          "value": 0.03744757339741005,
          "variable": "time"
         },
         {
          "max_depth": 4,
          "value": 0.24460754940270013,
          "variable": "time"
         },
         {
          "max_depth": 2,
          "value": 0.009858055207996852,
          "variable": "scoreTimeRatio"
         },
         {
          "max_depth": 9,
          "value": 0.1389296155836436,
          "variable": "scoreTimeRatio"
         },
         {
          "max_depth": 6,
          "value": 1,
          "variable": "scoreTimeRatio"
         },
         {
          "max_depth": 4,
          "value": 0,
          "variable": "scoreTimeRatio"
         }
        ]
       },
       "encoding": {
        "color": {
         "field": "variable",
         "type": "nominal"
        },
        "x": {
         "field": "max_depth",
         "type": "ordinal"
        },
        "y": {
         "field": "value",
         "type": "quantitative"
        }
       },
       "mark": "line",
       "width": 400
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 2 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "max_depth = bgs.compareValsBaseCase(X, \n",
    "                    y, \n",
    "                    getForestAccuracy,        \n",
    "                    rfArgs, \n",
    "                    roc_auc_score, \n",
    "                    \"max_depth\", \n",
    "                    0, \n",
    "                    2, \n",
    "                    9)\n",
    "bgs.showTimeScoreChartAndGraph(max_depth)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:py36]",
   "language": "python",
   "name": "conda-env-py36-py"
  },
  "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
}