10장_차원축소_4절_LDA.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 개별 코드 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 데이터 불러오기\n",
    "from sklearn import datasets\n",
    "raw_wine = datasets.load_wine()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 피쳐, 타겟 데이터 지정\n",
    "X = raw_wine.data\n",
    "y = raw_wine.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 트레이닝/테스트 데이터 분할\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_tn, X_te, y_tn, y_te=train_test_split(X,y,random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 데이터 표준화\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "std_scale = StandardScaler()\n",
    "std_scale.fit(X_tn)\n",
    "X_tn_std = std_scale.transform(X_tn)\n",
    "X_te_std  = std_scale.transform(X_te)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# LDA\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
    "lda = LinearDiscriminantAnalysis()\n",
    "lda.fit(X_tn_std, y_tn)\n",
    "X_tn_lda = lda.transform(X_tn_std)\n",
    "X_te_lda = lda.transform(X_te_std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(133, 13)\n",
      "(133, 2)\n"
     ]
    }
   ],
   "source": [
    "# 차원축소 전후 데이터 차원 비교\n",
    "print(X_tn_std.shape)\n",
    "print(X_tn_lda.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ -9.49063922  -3.6792656  -10.65310672]\n"
     ]
    }
   ],
   "source": [
    "# LDA 상수항\n",
    "print(lda.intercept_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 2.11536841  0.34171834  1.60756274 -2.62777005 -0.02954383 -1.92461007\n",
      "   5.43612614  0.30142527 -0.66723405 -2.08948728  0.53569108  3.06011656\n",
      "   5.45318503]\n",
      " [-1.18707715 -0.97635273 -1.64104916  1.19614378  0.22114803  0.2969019\n",
      "   1.13721801  0.4338713   0.33425086 -1.57541361  0.56544507 -0.17730553\n",
      "  -2.34856045]\n",
      " [-0.59547207  1.01875252  0.59754164  1.13544231 -0.28238675  1.65463974\n",
      "  -7.48134065 -0.94177595  0.24492236  4.49319245 -1.38150969 -3.04974421\n",
      "  -2.54627163]]\n"
     ]
    }
   ],
   "source": [
    "# LDA 가중 벡터\n",
    "print(lda.coef_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>lda_comp1</th>\n",
       "      <th>lda_comp2</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-2.563293</td>\n",
       "      <td>-0.503212</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.646015</td>\n",
       "      <td>-2.169761</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-3.981901</td>\n",
       "      <td>1.621584</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-5.956216</td>\n",
       "      <td>3.878140</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-4.047839</td>\n",
       "      <td>2.335699</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   lda_comp1  lda_comp2  target\n",
       "0  -2.563293  -0.503212       0\n",
       "1  -0.646015  -2.169761       1\n",
       "2  -3.981901   1.621584       0\n",
       "3  -5.956216   3.878140       0\n",
       "4  -4.047839   2.335699       0"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# LDA 데이터 셋\n",
    "import pandas as pd\n",
    "lda_columns = ['lda_comp1', 'lda_comp2']\n",
    "X_tn_lda_df = pd.DataFrame(X_tn_lda, \n",
    "                           columns=lda_columns)\n",
    "X_tn_lda_df['target'] = y_tn\n",
    "X_tn_lda_df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "# LDA 시각화\n",
    "import matplotlib.pyplot as plt\n",
    "df = X_tn_lda_df\n",
    "markers=['o','x','^']\n",
    "\n",
    "for i, mark in enumerate(markers):\n",
    "    X_i = df[df['target']== i]\n",
    "    target_i = raw_wine.target_names[i]\n",
    "    X1 = X_i['lda_comp1']\n",
    "    X2 = X_i['lda_comp2']\n",
    "    plt.scatter(X1, X2, \n",
    "                marker=mark, \n",
    "                label=target_i)\n",
    "plt.xlabel('lda_component1')\n",
    "plt.ylabel('lda_component2')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "# 학습\n",
    "clf_rf_lda = RandomForestClassifier(max_depth=2, \n",
    "                                    random_state=0)\n",
    "clf_rf_lda.fit(X_tn_lda, y_tn)\n",
    "\n",
    "# 예측\n",
    "pred_rf_lda = clf_rf_lda.predict(X_te_lda)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n"
     ]
    }
   ],
   "source": [
    "# LDA 적용 후 정확도\n",
    "from sklearn.metrics import accuracy_score\n",
    "accuracy_lda = accuracy_score(y_te, pred_rf_lda)\n",
    "print(accuracy_lda)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 통합 코드"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(133, 13)\n",
      "(133, 2)\n",
      "[ -9.49063922  -3.6792656  -10.65310672]\n",
      "[[ 2.11536841  0.34171834  1.60756274 -2.62777005 -0.02954383 -1.92461007\n",
      "   5.43612614  0.30142527 -0.66723405 -2.08948728  0.53569108  3.06011656\n",
      "   5.45318503]\n",
      " [-1.18707715 -0.97635273 -1.64104916  1.19614378  0.22114803  0.2969019\n",
      "   1.13721801  0.4338713   0.33425086 -1.57541361  0.56544507 -0.17730553\n",
      "  -2.34856045]\n",
      " [-0.59547207  1.01875252  0.59754164  1.13544231 -0.28238675  1.65463974\n",
      "  -7.48134065 -0.94177595  0.24492236  4.49319245 -1.38150969 -3.04974421\n",
      "  -2.54627163]]\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n"
     ]
    }
   ],
   "source": [
    "from sklearn import datasets\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
    "\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "# 데이터 불러오기\n",
    "raw_wine = datasets.load_wine()\n",
    "\n",
    "# 피쳐, 타겟 데이터 지정\n",
    "X = raw_wine.data\n",
    "y = raw_wine.target\n",
    "\n",
    "# 트레이닝/테스트 데이터 분할\n",
    "X_tn, X_te, y_tn, y_te=train_test_split(X,y,random_state=1)\n",
    "\n",
    "# 데이터 표준화\n",
    "std_scale = StandardScaler()\n",
    "std_scale.fit(X_tn)\n",
    "X_tn_std = std_scale.transform(X_tn)\n",
    "X_te_std  = std_scale.transform(X_te)\n",
    "\n",
    "# LDA\n",
    "lda = LinearDiscriminantAnalysis()\n",
    "lda.fit(X_tn_std, y_tn)\n",
    "X_tn_lda = lda.transform(X_tn_std)\n",
    "X_te_lda = lda.transform(X_te_std)\n",
    "\n",
    "# 차원축소 전후 데이터 차원 비교\n",
    "print(X_tn_std.shape)\n",
    "print(X_tn_lda.shape)\n",
    "\n",
    "# LDA 상수항\n",
    "print(lda.intercept_)\n",
    "\n",
    "# LDA 가중 벡터\n",
    "print(lda.coef_)\n",
    "\n",
    "# LDA 적용 된 데이터프레임\n",
    "lda_columns = ['lda_comp1', 'lda_comp2']\n",
    "X_tn_lda_df = pd.DataFrame(X_tn_lda, \n",
    "                           columns=lda_columns)\n",
    "X_tn_lda_df['target'] = y_tn\n",
    "X_tn_lda_df.head(5)\n",
    "\n",
    "# LDA 시각화\n",
    "df = X_tn_lda_df\n",
    "markers=['o','x','^']\n",
    "\n",
    "for i, mark in enumerate(markers):\n",
    "    X_i = df[df['target']== i]\n",
    "    target_i = raw_wine.target_names[i]\n",
    "    X1 = X_i['lda_comp1']\n",
    "    X2 = X_i['lda_comp2']\n",
    "    plt.scatter(X1, X2, \n",
    "                marker=mark, \n",
    "                label=target_i)\n",
    "plt.xlabel('lda_component1')\n",
    "plt.ylabel('lda_component2')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# 학습\n",
    "clf_rf_lda = RandomForestClassifier(max_depth=2, \n",
    "                                    random_state=0)\n",
    "clf_rf_lda.fit(X_tn_lda, y_tn)\n",
    "\n",
    "# 예측\n",
    "pred_rf_lda = clf_rf_lda.predict(X_te_lda)\n",
    "\n",
    "# LDA 적용 후 정확도\n",
    "accuracy_lda = accuracy_score(y_te, pred_rf_lda)\n",
    "print(accuracy_lda)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}