how_feature_importance_is_calculated.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "01ee91aa",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import make_classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "5d938347",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b2445e3e",
   "metadata": {},
   "outputs": [],
   "source": [
    "X,y=make_classification(n_samples=5,n_classes=2,n_features=2,n_informative=2,n_redundant=0,random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "d89502be",
   "metadata": {},
   "outputs": [],
   "source": [
    "clf=DecisionTreeClassifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8d80e9f0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeClassifier()"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d85c7f26",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(133.92000000000002, 181.2, 'X[1] <= -0.894\\ngini = 0.48\\nsamples = 5\\nvalue = [3, 2]'),\n",
       " Text(66.96000000000001, 108.72, 'gini = 0.0\\nsamples = 1\\nvalue = [0, 1]'),\n",
       " Text(200.88000000000002, 108.72, 'X[0] <= 1.01\\ngini = 0.375\\nsamples = 4\\nvalue = [3, 1]'),\n",
       " Text(133.92000000000002, 36.23999999999998, 'gini = 0.0\\nsamples = 3\\nvalue = [3, 0]'),\n",
       " Text(267.84000000000003, 36.23999999999998, 'gini = 0.0\\nsamples = 1\\nvalue = [0, 1]')]"
      ]
     },
     "execution_count": 6,
     "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": [
    "from sklearn.tree import plot_tree\n",
    "plot_tree(clf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "da5787af",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.625, 0.375])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "3750b8c9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(n_estimators=2)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "rf = RandomForestClassifier(n_estimators=2)\n",
    "rf.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "6976540b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1., 0.])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "cecc8cc4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 0.]\n",
      "[1. 0.]\n",
      "0.7775000000000001\n"
     ]
    }
   ],
   "source": [
    "print(rf.estimators_[0].feature_importances_)\n",
    "print(rf.estimators_[1].feature_importances_)\n",
    "\n",
    "print((1 + 0.555)/2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "4769e07d",
   "metadata": {},
   "outputs": [],
   "source": [
    "X,y = make_classification(n_samples=15, n_classes=2,\n",
    "                               n_features=2, n_informative=2, n_redundant=0,\n",
    "                               random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "4924d6eb",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "991a4318",
   "metadata": {},
   "outputs": [],
   "source": [
    "clf = DecisionTreeClassifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "6d7a5d6b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeClassifier()"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "c2f203bf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(111.60000000000001, 190.26, 'X[1] <= -0.383\\ngini = 0.498\\nsamples = 15\\nvalue = [8, 7]'),\n",
       " Text(55.800000000000004, 135.9, 'gini = 0.0\\nsamples = 6\\nvalue = [6, 0]'),\n",
       " Text(167.4, 135.9, 'X[0] <= 0.878\\ngini = 0.346\\nsamples = 9\\nvalue = [2, 7]'),\n",
       " Text(111.60000000000001, 81.53999999999999, 'gini = 0.0\\nsamples = 6\\nvalue = [0, 6]'),\n",
       " Text(223.20000000000002, 81.53999999999999, 'X[1] <= -0.278\\ngini = 0.444\\nsamples = 3\\nvalue = [2, 1]'),\n",
       " Text(167.4, 27.180000000000007, 'gini = 0.0\\nsamples = 1\\nvalue = [0, 1]'),\n",
       " Text(279.0, 27.180000000000007, 'gini = 0.0\\nsamples = 2\\nvalue = [2, 0]')]"
      ]
     },
     "execution_count": 15,
     "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": [
    "from sklearn.tree import plot_tree\n",
    "plot_tree(clf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "c98ad013",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.23809524, 0.76190476])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f46cc0fa",
   "metadata": {},
   "outputs": [],
   "source": [
    "clf."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "79986064",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(n_estimators=2)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "rf=RandomForestClassifier(n_estimators=2)\n",
    "rf.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "6460fa30",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.28571429 0.71428571]\n",
      "[0.76190476 0.23809524]\n"
     ]
    }
   ],
   "source": [
    "print(rf.estimators_[0].feature_importances_)\n",
    "print(rf.estimators_[1].feature_importances_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "c7874531",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.523809525"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(0.28571429 +0.76190476 )/2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "59ef9ed4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.476190475"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(0.71428571+0.23809524)/2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "c750bb31",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.52380952, 0.47619048])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0109f495",
   "metadata": {},
   "outputs": [],
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}