3_pca_short_ex.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "RtGPFVWP-xic"
   },
   "outputs": [],
   "source": [
    "data_dir = './data/'\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 286
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1036,
     "status": "ok",
     "timestamp": 1565666321216,
     "user": {
      "displayName": "竜馬備瀬",
      "photoUrl": "",
      "userId": "17932567482837919211"
     },
     "user_tz": -540
    },
    "id": "XWGX0PX4--e7",
    "outputId": "a16de6f7-2752-470c-d328-779f851ccc56"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x24c7406a348>"
      ]
     },
     "execution_count": 2,
     "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": [
    "####################################\n",
    "# 練習１\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Read the iris dataset\n",
    "iris = pd.read_csv(data_dir+'iris.csv')\n",
    "\n",
    "# get the \"petal width\" data\n",
    "X = iris['petal width']\n",
    "\n",
    "# データをプロット\n",
    "plt.scatter(range(X.size), X, color=\"steelblue\")\n",
    "# 正規化したデータをプロット\n",
    "X_zero_mean = X-X.mean()\n",
    "X_scaled = X_zero_mean/X_zero_mean.std()\n",
    "plt.scatter(range(X_scaled.size), X_scaled, color=\"orange\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 773
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1140,
     "status": "ok",
     "timestamp": 1565666387620,
     "user": {
      "displayName": "竜馬備瀬",
      "photoUrl": "",
      "userId": "17932567482837919211"
     },
     "user_tz": -540
    },
    "id": "_SbJ83L3_LUd",
    "outputId": "c777e14c-2afd-424f-b673-1a5be2715ef5"
   },
   "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"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "###############################\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Read the iris dataset\n",
    "iris = pd.read_csv(data_dir+'iris.csv')\n",
    "# get the \"petal width\" data\n",
    "X = iris['petal width']\n",
    "plt.hist(X)\n",
    "plt.show()\n",
    "\n",
    "# substract the mean to all data to obtain the centered data\n",
    "X_zero_mean = X-X.mean()\n",
    "plt.hist(X_zero_mean)\n",
    "plt.show()\n",
    "\n",
    "# Divide by the standard deviation to normalize the data\n",
    "X_scaled = X_zero_mean/X_zero_mean.std()\n",
    "plt.hist(X_scaled)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 357
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 873,
     "status": "ok",
     "timestamp": 1565666432058,
     "user": {
      "displayName": "竜馬備瀬",
      "photoUrl": "",
      "userId": "17932567482837919211"
     },
     "user_tz": -540
    },
    "id": "xPcdBT96_Xuv",
    "outputId": "67bb46c5-a020-4369-cc21-530fbdd7db6d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.72770452 0.23030523 0.03683832 0.00515193]\n",
      "[[ 0.52237162 -0.26335492  0.58125401  0.56561105]\n",
      " [ 0.37231836  0.92555649  0.02109478  0.06541577]\n",
      " [-0.72101681  0.24203288  0.14089226  0.6338014 ]\n",
      " [-0.26199559  0.12413481  0.80115427 -0.52354627]]\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"
    }
   ],
   "source": [
    "####################################\n",
    "# 練習2\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "\n",
    "X = iris.drop('class', axis=1)\n",
    "\n",
    "# Normalize the data\n",
    "X_scaled = (X-X.mean())/X.std()\n",
    "\n",
    "# Run PCA\n",
    "pca = PCA(n_components=4)\n",
    "pca.fit(X_scaled)\n",
    "\n",
    "# Plot the importance of each component\n",
    "plt.plot(pca.explained_variance_ratio_)\n",
    "\n",
    "print(pca.explained_variance_ratio_)\n",
    "print(pca.components_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 105
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 685,
     "status": "ok",
     "timestamp": 1565666451470,
     "user": {
      "displayName": "竜馬備瀬",
      "photoUrl": "",
      "userId": "17932567482837919211"
     },
     "user_tz": -540
    },
    "id": "nQQ0CzEt_g45",
    "outputId": "0590533e-cf2a-48c7-bcda-62c2cacd7729"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained variance ratio:  [0.72770452 0.23030523 0.03683832 0.00515193]\n",
      "the 4 principal vectors:  [[ 0.52237162 -0.26335492  0.58125401  0.56561105]\n",
      " [-0.37231836 -0.92555649 -0.02109478 -0.06541577]\n",
      " [-0.72101681  0.24203288  0.14089226  0.6338014 ]\n",
      " [ 0.26199559 -0.12413481 -0.80115427  0.52354627]]\n"
     ]
    }
   ],
   "source": [
    "####################################\n",
    "# 練習3\n",
    "\n",
    "X = iris.drop('class', axis=1)\n",
    "\n",
    "# Normalize the data\n",
    "X_scaled = (X-X.mean())/X.std()\n",
    "\n",
    "# Transform input data into an np.array\n",
    "Y = np.array(X_scaled)\n",
    "\n",
    "# Compute the covariance matrix\n",
    "cov_matrix = np.cov(Y.transpose())\n",
    "\n",
    "# Compute the eigen values and eigen vectors\n",
    "(l,V) = np.linalg.eig(cov_matrix)\n",
    "\n",
    "# you should obtain the same results as before, modulo the sign of the principal vectors\n",
    "print(\"Explained variance ratio: \", l/l.sum())\n",
    "print(\"the 4 principal vectors: \", V.transpose())\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "4H88h_FA_A5Z"
   },
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "name": "PCA_short_ex.ipynb",
   "provenance": [],
   "version": "0.3.2"
  },
  "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
}