PCA.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "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>food</th>\n",
       "      <th>Protein</th>\n",
       "      <th>fat</th>\n",
       "      <th>Vitamin C</th>\n",
       "      <th>Fiber</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Beef</td>\n",
       "      <td>19.42</td>\n",
       "      <td>12.73</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Borage</td>\n",
       "      <td>1.80</td>\n",
       "      <td>0.70</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Cabbage</td>\n",
       "      <td>1.28</td>\n",
       "      <td>0.10</td>\n",
       "      <td>36.6</td>\n",
       "      <td>2.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Chard</td>\n",
       "      <td>1.80</td>\n",
       "      <td>0.20</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Chicken</td>\n",
       "      <td>17.44</td>\n",
       "      <td>8.10</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      food  Protein    fat  Vitamin C  Fiber\n",
       "0     Beef    19.42  12.73        0.0    0.0\n",
       "1   Borage     1.80   0.70       35.0    0.0\n",
       "2  Cabbage     1.28   0.10       36.6    2.5\n",
       "3    Chard     1.80   0.20       30.0    1.6\n",
       "4  Chicken    17.44   8.10        0.0    0.0"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 讀取食物營養成分資料\n",
    "food = pd.read_csv('food_table.csv')\n",
    "food.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PCA(copy=True, iterated_power='auto', n_components=4, random_state=None,\n",
       "  svd_solver='auto', tol=0.0, whiten=False)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 將dataframe轉成numpy array\n",
    "food_array = food.values[:,1:5].astype(float)\n",
    "\n",
    "# 標準化\n",
    "std = StandardScaler()\n",
    "food_number_std = std.fit_transform(food_array)\n",
    "\n",
    "# PCA\n",
    "pca = PCA(n_components=4)\n",
    "pca.fit(food_number_std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>content</th>\n",
       "      <th>pc1</th>\n",
       "      <th>pc2</th>\n",
       "      <th>pc3</th>\n",
       "      <th>pc4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Protein</td>\n",
       "      <td>0.535869</td>\n",
       "      <td>0.371633</td>\n",
       "      <td>-0.029084</td>\n",
       "      <td>-0.757554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>fat</td>\n",
       "      <td>0.502184</td>\n",
       "      <td>0.580700</td>\n",
       "      <td>-0.013782</td>\n",
       "      <td>0.640631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Vitamin C</td>\n",
       "      <td>-0.483022</td>\n",
       "      <td>0.484500</td>\n",
       "      <td>-0.725363</td>\n",
       "      <td>-0.076145</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Fiber</td>\n",
       "      <td>-0.476808</td>\n",
       "      <td>0.538456</td>\n",
       "      <td>0.687614</td>\n",
       "      <td>-0.099527</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     content       pc1       pc2       pc3       pc4\n",
       "0    Protein  0.535869  0.371633 -0.029084 -0.757554\n",
       "1        fat  0.502184  0.580700 -0.013782  0.640631\n",
       "2  Vitamin C -0.483022  0.484500 -0.725363 -0.076145\n",
       "3      Fiber -0.476808  0.538456  0.687614 -0.099527"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca_w = pd.DataFrame()\n",
    "pca_w['content'] = food.columns[1:]\n",
    "pca_w['pc1'] = pca.components_[0]\n",
    "pca_w['pc2'] = pca.components_[1]\n",
    "pca_w['pc3'] = pca.components_[2]\n",
    "pca_w['pc4'] = pca.components_[3]\n",
    "pca_w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 變異解釋百分比\n",
    "ratio = pca.explained_variance_ratio_\n",
    "plt.style.use('seaborn-bright')\n",
    "fig = plt.figure(figsize=(10,6))\n",
    "ax = fig.add_subplot(111)\n",
    "\n",
    "# 設置x, y軸\n",
    "xx = [1, 2, 3, 4]\n",
    "x_trick = ['PC1', 'PC2', 'PC3', 'PC4',]\n",
    "ax.set_xticks(xx)\n",
    "ax.set_xticklabels(x_trick)\n",
    "\n",
    "plt.plot(xx, ratio, c='#AAAAAA', lw=3)\n",
    "plt.scatter(xx, ratio, marker='o', c='#AAAAAA', s=120)\n",
    "\n",
    "# 設置坐標軸的字體大小\n",
    "plt.rc('xtick', labelsize=18)\n",
    "plt.rc('ytick', labelsize=18)\n",
    "\n",
    "# 坐標軸名稱\n",
    "plt.xlabel('Number of Components', fontsize=20)\n",
    "plt.ylabel('% of Data Spread Accounted', fontsize=15)\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}