2.Data preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " データ前処理\n",
    "\n",
    "1. 欠損値処理\n",
    "2. 外れ値検出と処理\n",
    "3. データの標準化・正規化\n",
    "4. データの型変換\n",
    "5. テキストデータの前処理\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "元のデータ:\n",
      "      A    B\n",
      "0  1.0  NaN\n",
      "1  2.0  2.0\n",
      "2  NaN  3.0\n",
      "3  4.0  4.0\n",
      "4  5.0  NaN\n",
      "平均値補完:\n",
      "      A    B\n",
      "0  1.0  3.0\n",
      "1  2.0  2.0\n",
      "2  3.0  3.0\n",
      "3  4.0  4.0\n",
      "4  5.0  3.0\n",
      "前のデータで補完:\n",
      "      A    B\n",
      "0  1.0  NaN\n",
      "1  2.0  2.0\n",
      "2  2.0  3.0\n",
      "3  4.0  4.0\n",
      "4  5.0  4.0\n",
      "欠損値を含む行の削除:\n",
      "      A    B\n",
      "1  2.0  2.0\n",
      "3  4.0  4.0\n"
     ]
    }
   ],
   "source": [
    "# 1. 欠損値処理 (平均値補完、前後データ補完、削除)\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "# サンプルデータ作成\n",
    "data = {'A': [1, 2, np.nan, 4, 5],\n",
    "        'B': [np.nan, 2, 3, 4, np.nan]}\n",
    "df = pd.DataFrame(data)\n",
    "\n",
    "# 平均値補完\n",
    "df_mean = df.fillna(df.mean())\n",
    "# 前のデータで補完\n",
    "df_ffill = df.fillna(method='ffill')\n",
    "# 欠損値を含む行を削除\n",
    "df_drop = df.dropna()\n",
    "\n",
    "print('元のデータ:\\n', df)\n",
    "print('平均値補完:\\n', df_mean)\n",
    "print('前のデータで補完:\\n', df_ffill)\n",
    "print('欠損値を含む行の削除:\\n', df_drop)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "IQR法による外れ値除去:\n",
      "      A    B\n",
      "0  1.0  NaN\n",
      "1  2.0  2.0\n",
      "2  NaN  3.0\n",
      "3  4.0  4.0\n",
      "4  5.0  NaN\n",
      "Zスコアによる外れ値除去:\n",
      "      A    B\n",
      "0  1.0  NaN\n",
      "1  2.0  2.0\n",
      "2  NaN  3.0\n",
      "3  4.0  4.0\n",
      "4  5.0  NaN\n"
     ]
    }
   ],
   "source": [
    "# 2. 外れ値検出と処理 (IQR法, Zスコア)\n",
    "from scipy import stats\n",
    "\n",
    "# IQR法による外れ値検出\n",
    "Q1 = df.quantile(0.25)\n",
    "Q3 = df.quantile(0.75)\n",
    "IQR = Q3 - Q1\n",
    "df_no_outliers = df[~((df < (Q1 - 1.5 * IQR)) | (df > (Q3 + 1.5 * IQR))).any(axis=1)]\n",
    "\n",
    "# Zスコアによる外れ値検出\n",
    "z_scores = np.abs(stats.zscore(df.fillna(0)))\n",
    "df_zscore = df[(z_scores < 3).all(axis=1)]\n",
    "\n",
    "print('IQR法による外れ値除去:\\n', df_no_outliers)\n",
    "print('Zスコアによる外れ値除去:\\n', df_zscore)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Min-Maxスケーリング:\n",
      "      A     B\n",
      "0  0.2  0.00\n",
      "1  0.4  0.50\n",
      "2  0.0  0.75\n",
      "3  0.8  1.00\n",
      "4  1.0  0.00\n",
      "Zスコア正規化:\n",
      "           A      B\n",
      "0 -0.754829 -1.125\n",
      "1 -0.215666  0.125\n",
      "2 -1.293993  0.750\n",
      "3  0.862662  1.375\n",
      "4  1.401826 -1.125\n"
     ]
    }
   ],
   "source": [
    "# 3. データの標準化・正規化 (Min-Maxスケーリング, Zスコア正規化)\n",
    "from sklearn.preprocessing import MinMaxScaler, StandardScaler\n",
    "\n",
    "# Min-Maxスケーリング\n",
    "scaler_minmax = MinMaxScaler()\n",
    "df_minmax = pd.DataFrame(scaler_minmax.fit_transform(df.fillna(0)), columns=df.columns)\n",
    "\n",
    "# Zスコア正規化\n",
    "scaler_zscore = StandardScaler()\n",
    "df_zscore_scaled = pd.DataFrame(scaler_zscore.fit_transform(df.fillna(0)), columns=df.columns)\n",
    "\n",
    "print('Min-Maxスケーリング:\\n', df_minmax)\n",
    "print('Zスコア正規化:\\n', df_zscore_scaled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "元のカテゴリデータ:\n",
      "    Color\n",
      "0    Red\n",
      "1   Blue\n",
      "2  Green\n",
      "3   Blue\n",
      "4    Red\n",
      "One-Hotエンコーディング後:\n",
      "    Color_Blue  Color_Green  Color_Red\n",
      "0       False        False       True\n",
      "1        True        False      False\n",
      "2       False         True      False\n",
      "3        True        False      False\n",
      "4       False        False       True\n"
     ]
    }
   ],
   "source": [
    "# 4. データの型変換 (カテゴリデータのOne-Hotエンコーディング)\n",
    "df_cat = pd.DataFrame({'Color': ['Red', 'Blue', 'Green', 'Blue', 'Red']})\n",
    "\n",
    "# One-Hotエンコーディング\n",
    "df_encoded = pd.get_dummies(df_cat)\n",
    "\n",
    "print('元のカテゴリデータ:\\n', df_cat)\n",
    "print('One-Hotエンコーディング後:\\n', df_encoded)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.decomposition import PCA\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import rcParams\n",
    "\n",
    "rcParams['font.family'] = 'Meiryo'\n",
    "#  主成分分析（PCA）\n",
    "# データの生成\n",
    "X_pca = np.random.rand(100, 5) * 10  # 5次元データ\n",
    "scaler = StandardScaler()\n",
    "X_scaled = scaler.fit_transform(X_pca)\n",
    "\n",
    "# PCAの適用\n",
    "pca = PCA(n_components=2)\n",
    "X_pca_2d = pca.fit_transform(X_scaled)\n",
    "\n",
    "# 可視化\n",
    "plt.figure(figsize=(6, 4))\n",
    "plt.scatter(X_pca_2d[:, 0], X_pca_2d[:, 1], alpha=0.7)\n",
    "plt.title(\"PCA による次元削減\")\n",
    "plt.xlabel(\"主成分1\")\n",
    "plt.ylabel(\"主成分2\")\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "dataSci",
   "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": "3.8.20"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}