PCA.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n",
      " 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2\n",
      " 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2\n",
      " 2 2]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.datasets import load_iris\n",
    "from sklearn.decomposition import PCA\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "iris = load_iris()\n",
    "X = iris.data\n",
    "y = iris.target\n",
    "print(y)\n",
    "m = X.shape[0]\n",
    "n = X.shape[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     principal component 1\n",
      "0                -2.684126\n",
      "1                -2.714142\n",
      "2                -2.888991\n",
      "3                -2.745343\n",
      "4                -2.728717\n",
      "..                     ...\n",
      "145               1.944110\n",
      "146               1.527167\n",
      "147               1.764346\n",
      "148               1.900942\n",
      "149               1.390189\n",
      "\n",
      "[150 rows x 1 columns]\n",
      "[0.92461872]\n"
     ]
    }
   ],
   "source": [
    "# Carry out a principal component analysis on the entire rawdataset k=1 component.\n",
    "pca = PCA(n_components=1)\n",
    "principalComponents = pca.fit_transform(X)\n",
    "principalDf = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['principal component 1'])\n",
    "print(principalDf)\n",
    "print(pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     principal component 1  principal component 2\n",
      "0                -2.684126               0.319397\n",
      "1                -2.714142              -0.177001\n",
      "2                -2.888991              -0.144949\n",
      "3                -2.745343              -0.318299\n",
      "4                -2.728717               0.326755\n",
      "..                     ...                    ...\n",
      "145               1.944110               0.187532\n",
      "146               1.527167              -0.375317\n",
      "147               1.764346               0.078859\n",
      "148               1.900942               0.116628\n",
      "149               1.390189              -0.282661\n",
      "\n",
      "[150 rows x 2 columns]\n",
      "[0.92461872 0.05306648]\n"
     ]
    }
   ],
   "source": [
    "# Carry out a principal component analysis on the entire rawdataset k=2 components.\n",
    "pca = PCA(n_components=2)\n",
    "principalComponents = pca.fit_transform(X)\n",
    "principalDf = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['principal component 1', 'principal component 2'])\n",
    "print(principalDf)\n",
    "print(pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     principal component 1  principal component 2  principal component 3\n",
      "0                -2.684126               0.319397              -0.027915\n",
      "1                -2.714142              -0.177001              -0.210464\n",
      "2                -2.888991              -0.144949               0.017900\n",
      "3                -2.745343              -0.318299               0.031559\n",
      "4                -2.728717               0.326755               0.090079\n",
      "..                     ...                    ...                    ...\n",
      "145               1.944110               0.187532               0.177825\n",
      "146               1.527167              -0.375317              -0.121898\n",
      "147               1.764346               0.078859               0.130482\n",
      "148               1.900942               0.116628               0.723252\n",
      "149               1.390189              -0.282661               0.362910\n",
      "\n",
      "[150 rows x 3 columns]\n",
      "[0.92461872 0.05306648 0.01710261]\n"
     ]
    }
   ],
   "source": [
    "# Carry out a principal component analysis on the entire rawdataset k=3 components.\n",
    "pca = PCA(n_components=3)\n",
    "principalComponents = pca.fit_transform(X)\n",
    "principalDf = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['principal component 1', 'principal component 2', 'principal component 3'])\n",
    "print(principalDf)\n",
    "print(pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     principal component 1  principal component 2  principal component 3  \\\n",
      "0                -2.684126               0.319397              -0.027915   \n",
      "1                -2.714142              -0.177001              -0.210464   \n",
      "2                -2.888991              -0.144949               0.017900   \n",
      "3                -2.745343              -0.318299               0.031559   \n",
      "4                -2.728717               0.326755               0.090079   \n",
      "..                     ...                    ...                    ...   \n",
      "145               1.944110               0.187532               0.177825   \n",
      "146               1.527167              -0.375317              -0.121898   \n",
      "147               1.764346               0.078859               0.130482   \n",
      "148               1.900942               0.116628               0.723252   \n",
      "149               1.390189              -0.282661               0.362910   \n",
      "\n",
      "     principal component 4  \n",
      "0                -0.002262  \n",
      "1                -0.099027  \n",
      "2                -0.019968  \n",
      "3                 0.075576  \n",
      "4                 0.061259  \n",
      "..                     ...  \n",
      "145              -0.426196  \n",
      "146              -0.254367  \n",
      "147              -0.137001  \n",
      "148              -0.044595  \n",
      "149               0.155039  \n",
      "\n",
      "[150 rows x 4 columns]\n",
      "[0.92461872 0.05306648 0.01710261 0.00521218]\n"
     ]
    }
   ],
   "source": [
    "# Carry out a principal component analysis on the entire rawdataset k=4 components.\n",
    "pca = PCA(n_components=4)\n",
    "principalComponents = pca.fit_transform(X)\n",
    "principalDf = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['principal component 1', 'principal component 2', 'principal component 3', 'principal component 4'])\n",
    "print(principalDf)\n",
    "print(pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "How much of variance in data can be explained by the first principal component?\n",
    "After calling the pce.explained_variance_ratio_ method, we can see that the first principal component accounts for \n",
    "92.461872% of the variance in the data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "How does the fraction of variation explained in data vary askvaries?\n",
    "As k, the number of principal components increases, the percentage of information observed from the data increases as\n",
    "well: k=1: 92.461872%, k=2: 97.76852%, k=3: 99.478781%, k=4: 99.999999%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[5.84333333 3.05733333 3.758      1.19933333]\n"
     ]
    }
   ],
   "source": [
    "#(a)  Centering or mean removal and (c) Standardization\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "col_mean = np.zeros(n)\n",
    "\n",
    "for j in range(0, n): \n",
    "    for i in range(0, m):\n",
    "        col_mean[j] += X[i][j]\n",
    "        \n",
    "final_array = col_mean/m\n",
    "print(final_array)\n",
    "\n",
    "for j in range(0, n): \n",
    "    for i in range(0, m):\n",
    "        X[i][j] -= final_array[j]\n",
    "\n",
    "X = StandardScaler().fit_transform(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     principal component 1\n",
      "0                -2.264703\n",
      "1                -2.080961\n",
      "2                -2.364229\n",
      "3                -2.299384\n",
      "4                -2.389842\n",
      "..                     ...\n",
      "145               1.870503\n",
      "146               1.564580\n",
      "147               1.521170\n",
      "148               1.372788\n",
      "149               0.960656\n",
      "\n",
      "[150 rows x 1 columns]\n",
      "[0.72962445]\n"
     ]
    }
   ],
   "source": [
    "# Carry out a principal component analysis on the entire rawdataset k=1 component.\n",
    "pca = PCA(n_components=1)\n",
    "principalComponents = pca.fit_transform(X)\n",
    "principalDf = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['principal component 1'])\n",
    "print(principalDf)\n",
    "print(pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     principal component 1  principal component 2\n",
      "0                -2.264703               0.480027\n",
      "1                -2.080961              -0.674134\n",
      "2                -2.364229              -0.341908\n",
      "3                -2.299384              -0.597395\n",
      "4                -2.389842               0.646835\n",
      "..                     ...                    ...\n",
      "145               1.870503               0.386966\n",
      "146               1.564580              -0.896687\n",
      "147               1.521170               0.269069\n",
      "148               1.372788               1.011254\n",
      "149               0.960656              -0.024332\n",
      "\n",
      "[150 rows x 2 columns]\n",
      "[0.72962445 0.22850762]\n"
     ]
    }
   ],
   "source": [
    "# Carry out a principal component analysis on the entire rawdataset k=2 components.\n",
    "pca = PCA(n_components=2)\n",
    "principalComponents = pca.fit_transform(X)\n",
    "principalDf = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['principal component 1', 'principal component 2'])\n",
    "print(principalDf)\n",
    "print(pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     principal component 1  principal component 2  principal component 3\n",
      "0                -2.264703               0.480027              -0.127706\n",
      "1                -2.080961              -0.674134              -0.234609\n",
      "2                -2.364229              -0.341908               0.044201\n",
      "3                -2.299384              -0.597395               0.091290\n",
      "4                -2.389842               0.646835               0.015738\n",
      "..                     ...                    ...                    ...\n",
      "145               1.870503               0.386966               0.256274\n",
      "146               1.564580              -0.896687              -0.026371\n",
      "147               1.521170               0.269069               0.180178\n",
      "148               1.372788               1.011254               0.933395\n",
      "149               0.960656              -0.024332               0.528249\n",
      "\n",
      "[150 rows x 3 columns]\n",
      "[0.72962445 0.22850762 0.03668922]\n"
     ]
    }
   ],
   "source": [
    "# Carry out a principal component analysis on the entire rawdataset k=3 components.\n",
    "pca = PCA(n_components=3)\n",
    "principalComponents = pca.fit_transform(X)\n",
    "principalDf = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['principal component 1', 'principal component 2', 'principal component 3'])\n",
    "print(principalDf)\n",
    "print(pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     principal component 1  principal component 2  principal component 3  \\\n",
      "0                -2.264703               0.480027              -0.127706   \n",
      "1                -2.080961              -0.674134              -0.234609   \n",
      "2                -2.364229              -0.341908               0.044201   \n",
      "3                -2.299384              -0.597395               0.091290   \n",
      "4                -2.389842               0.646835               0.015738   \n",
      "..                     ...                    ...                    ...   \n",
      "145               1.870503               0.386966               0.256274   \n",
      "146               1.564580              -0.896687              -0.026371   \n",
      "147               1.521170               0.269069               0.180178   \n",
      "148               1.372788               1.011254               0.933395   \n",
      "149               0.960656              -0.024332               0.528249   \n",
      "\n",
      "     principal component 4  \n",
      "0                -0.024168  \n",
      "1                -0.103007  \n",
      "2                -0.028377  \n",
      "3                 0.065956  \n",
      "4                 0.035923  \n",
      "..                     ...  \n",
      "145              -0.389257  \n",
      "146              -0.220192  \n",
      "147              -0.119171  \n",
      "148              -0.026129  \n",
      "149               0.163078  \n",
      "\n",
      "[150 rows x 4 columns]\n",
      "[0.72962445 0.22850762 0.03668922 0.00517871]\n"
     ]
    }
   ],
   "source": [
    "# Carry out a principal component analysis on the entire rawdataset k=4 components.\n",
    "pca = PCA(n_components=4)\n",
    "principalComponents = pca.fit_transform(X)\n",
    "principalDf = pd.DataFrame(data = principalComponents\n",
    "             , columns = ['principal component 1', 'principal component 2', 'principal component 3', 'principal component 4'])\n",
    "print(principalDf)\n",
    "print(pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Explain any differences you observe compared to part (a) and justify.\n",
    " \n",
    "In comparison to the results from part (a), the values within the principal components from k=1 to k=4 change as well as the percentage of variance observed by each component changes as well. For starters, when looking at principal component 1 from part (a), the first new values are:\n",
    "\n",
    "0                -2.684126\n",
    "1                -2.714142\n",
    "2                -2.888991\n",
    "3                -2.745343\n",
    "4                -2.728717\n",
    "\n",
    "while for principal component 1 from part (c), the first few values are:\n",
    "\n",
    "0                -2.264703\n",
    "1                -2.080961\n",
    "2                -2.364229\n",
    "3                -2.299384\n",
    "4                -2.389842\n",
    "\n",
    "Clearly, we can see that the component values will continue to be slightly different as k is increased. Along with these numbers being different after preprocessing through mean removal and Standardization, the percentage of variance drastically changes. The first component no longer holds as much of the majority of the variance as before. For part (a) we had:\n",
    "\n",
    "Principal Component 1 % of Variance: 0.92461872\n",
    "Principal Component 2 % of Variance: 0.05306648\n",
    "Principal Component 3 % of Variance: 0.01710261\n",
    "Principal Component 4 % of Variance: 0.00521218\n",
    "\n",
    "Totals for each k-value on variance:\n",
    "k=1: 92.461872%, k=2: 97.76852%, k=3: 99.478781%, k=4: 99.999999%\n",
    "\n",
    "while for part (c) we had:\n",
    "\n",
    "Principal Component 1 % of Variance: 0.72962445\n",
    "Principal Component 2 % of Variance: 0.22850762\n",
    "Principal Component 3 % of Variance: 0.03668922\n",
    "Principal Component 4 % of Variance: 0.00517871\n",
    "\n",
    "Totals for each k-value on variance:\n",
    "k=1: 72.962445%, k=2: 95.813207%, k=3: 99.482129%, k=4: 100%\n",
    "\n",
    "\n",
    "Justification-\n",
    "The general idea of the PCA algorithm is to reduce variance which is accomplished when the data is preprocessed to newer data with less dissimilarities with the rest. By having less percentage of data being represented by the first principal component, more can be represented by k=3 components when preprocessed data is used. Without preprocessing, 99.999999% of the data is represented by k=4 principal components, while with preprocessing a complete 100% is represented.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Project the raw four dimensional data down to a two dimensional subspace generated by first two top principle\n",
    "# components (PCs) from part (b) and produce a scatter plot of the data.\n",
    "\n",
    "import pandas as pd\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "finals = []\n",
    "for i in y:\n",
    "    if i == 0:\n",
    "        finals.append('Iris-setosa')\n",
    "    if i == 1:\n",
    "        finals.append('Iris-versicolor')\n",
    "    if i == 2:\n",
    "        finals.append('Iris-virginica') \n",
    "\n",
    "finals = pd.DataFrame(finals, columns=['Target'])\n",
    "finalDf = pd.concat([principalDf, finals], axis = 1)\n",
    "\n",
    "fig = plt.figure(figsize = (8,8))\n",
    "targets = ['Iris-setosa', 'Iris-versicolor', 'Iris-virginica']\n",
    "colors = ['g', 'y', 'c']\n",
    "\n",
    "ax = fig.add_subplot(1,1,1) \n",
    "ax.set_xlabel('Principal Component 1', fontsize = 15)\n",
    "ax.set_ylabel('Principal Component 2', fontsize = 15)\n",
    "ax.set_title('2 component PCA', fontsize = 20)\n",
    "\n",
    "\n",
    "for target, color in zip(targets,colors):\n",
    "    indicesToKeep = finalDf['Target'] == target\n",
    "    ax.scatter(finalDf.loc[indicesToKeep, 'principal component 1']\n",
    "               , finalDf.loc[indicesToKeep, 'principal component 2']\n",
    "               , c = color\n",
    "               , s = 50)\n",
    "\n",
    "ax.legend(targets)\n",
    "ax.grid()\n"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "From the scatter plot above, we can clearly see three different colored clusters for each label: Iris-setosa, Iris-versicolor, Iris-virginica."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Either use your k-means++ implementation from previous homework or from scikit-learn to cluster data from part (c)\n",
    "# into three clusters. \n",
    "\n",
    "from sklearn.cluster import KMeans\n",
    "import pandas as pd\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "finals = []\n",
    "for i in y:\n",
    "    if i == 0:\n",
    "        finals.append('Iris-setosa')\n",
    "    if i == 1:\n",
    "        finals.append('Iris-versicolor')\n",
    "    if i == 2:\n",
    "        finals.append('Iris-virginica') \n",
    "\n",
    "finals = pd.DataFrame(finals, columns=['Target'])\n",
    "finalDf = pd.concat([principalDf, finals], axis = 1)\n",
    "\n",
    "fig = plt.figure(figsize = (8,8))\n",
    "targets = ['Iris-setosa', 'Iris-versicolor', 'Iris-virginica']\n",
    "colors = ['g', 'y', 'c']\n",
    "\n",
    "ax = fig.add_subplot(1,1,1) \n",
    "ax.set_xlabel('Principal Component 1', fontsize = 15)\n",
    "ax.set_ylabel('Principal Component 2', fontsize = 15)\n",
    "ax.set_title('2 component PCA', fontsize = 20)\n",
    "km = KMeans(n_clusters=3, init='k-means++', max_iter=400).fit(X)\n",
    "\"\"\"\"\n",
    "for target, color in zip(targets,colors):\n",
    "    indicesToKeep = finalDf['Target'] == target\n",
    "    ax.scatter(finalDf.loc[indicesToKeep, 'principal component 1']\n",
    "               , finalDf.loc[indicesToKeep, 'principal component 2']\n",
    "               , c = color\n",
    "               , s = 50)\n",
    "\"\"\"\n",
    "y_km = km.fit_predict(X)\n",
    "# Plot the centers\n",
    "#plt.scatter(X[:, 0], X[:, 1], c=km.labels)\n",
    "plt.scatter(X[y_km==0, 0], X[y_km==0, 1], s=100, c='red', label ='Cluster 1')\n",
    "plt.scatter(X[y_km==1, 0], X[y_km==1, 1], s=100, c='blue', label ='Cluster 2')\n",
    "plt.scatter(X[y_km==2, 0], X[y_km==2, 1], s=100, c='green', label ='Cluster 3')\n",
    "ax.scatter(km.cluster_centers_[:, 0], km.cluster_centers_[:, 1], s=10, c='black', label = 'Centroids')\n",
    "ax.legend(targets)\n",
    "ax.grid()"
   ]
  },
  {
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}