{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### <p style=\"text-align: right;\"> &#9989; **Put your name here** </p>\n",
    "#### <p style=\"text-align: right;\"> &#9989; Put your group member names here</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Day 12 In-Class Assignment: Principal Component Analysis (PCA)\n",
    "\n",
    "<img src=\"http://lazyprogrammer.me/wp-content/uploads/2015/11/PCA.jpg\" width=\"400\"><a href=\"http://lazyprogrammer.me/tutorial-principal-components-analysis-pca//\"><p style=\"text-align: right;\">\n",
    "Image from http://lazyprogrammer.me/\n",
    "</p></a>\n",
    "\n",
    "PCA Explained Visually: [http://setosa.io/ev/principal-component-analysis/](http://setosa.io/ev/principal-component-analysis/)\n",
    "\n",
    "\n",
    "1. **Scientific motivation** \n",
    "    - Biometrics (Pattern Recognition)\n",
    "2. **Modeling tools** \n",
    "    - Principal Component Analysis (PCA) for feature selection\n",
    "3. **Programming concepts** \n",
    "    - Code Review\n",
    "4. **Python Programming Concepts** \n",
    "    - Saving objects using the pickle library\n",
    "    \n",
    "\n",
    "\n",
    "# Problem Description\n",
    "In Machine Learning, a good \"rule of thumb\" is to have more than 10 times the number of training samples as dimensions in your feature vector.  So if your feature vector has 4 features, then you should have 40 samples for training (this doesn't count the samples for testing).\n",
    "\n",
    "This is also called the \"curse of dimensionality\" and can sometimes be counter intuitive.  Basically we want more dimensions to ensure that we have captured the patterns we want to learn in the model. However, too many dimensions requires more training data. This can be a big problem when you have high dimensional data. \n",
    "\n",
    "For example, the face images provided by sci-kit learn, which we visited briefly at the end of the last class period and are using today in class, have 50 rows and 37 columns; this results in a vector length of 1850.  By the Machine Learning rule of thumb, we should have 18,500 examples in our training set (we only have at most 1560 examples and that does not include the ones we have set aside for testing. Many of you saw what happened when we ignored this rule.\n",
    "\n",
    "In this assignment we are going to use Principal Component Analysis (PCA) to to reduce the size of our feature vector and only select features which are the \"most\" descriptive."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Agenda for today's class\n",
    "\n",
    "</p>\n",
    "\n",
    "1. Review of Face Recognition from last time\n",
    "1. Pre-class assignment review\n",
    "1. Feature Selection using Principal Component Analysis\n",
    "1. Picking the best Feature Vectors\n",
    "1. Eigenfaces (just for fun -- and knowledge!)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# imports for the day\n",
    "import numpy as np\n",
    "from sklearn.datasets import fetch_lfw_people\n",
    "import matplotlib.pylab as plt\n",
    "from ipywidgets import interact\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.svm import SVC\n",
    "import pickle\n",
    "import time\n",
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# 1. Review of Face Recognition from last time\n",
    "\n",
    "&#9989; For each of the following steps; run the code and discuss as a group what the code is doing.  As a class we will continue the discussion. \n",
    "\n",
    "First, lets download and view the data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Import the data from scikit-learn\n",
    "import numpy as np\n",
    "from sklearn.datasets import fetch_lfw_people\n",
    "\n",
    "sk_data = fetch_lfw_people(min_faces_per_person=50, resize=0.4)\n",
    "\n",
    "feature_vectors = sk_data.data\n",
    "class_labels = sk_data.target\n",
    "categories = sk_data.target_names\n",
    "n_samples, n_features = feature_vectors.shape\n",
    "N, h, w = sk_data.images.shape\n",
    "n_classes = len(categories)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bb270b956a40466491df74e50f5edab0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(IntSlider(value=779, description='i', max=1559), Output()), _dom_classes=('widget-intera…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def browse_images(images, labels, categories):\n",
    "    n = len(images)\n",
    "    def view_image(i):\n",
    "        plt.imshow(images[i], cmap=plt.cm.gray, interpolation='nearest')\n",
    "        plt.title('%s' % categories[labels[i]])\n",
    "        plt.axis('off')\n",
    "        plt.show()\n",
    "    interact(view_image, i=(0,n-1))\n",
    "browse_images(sk_data.images, sk_data.target, sk_data.target_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 1: Splitting the dataset into training and testing sets\n",
    "\n",
    "Make train and test datasets as before."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# your answer here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 2: Train an SVM Classifier based on the training dataset.\n",
    "\n",
    "As we learned in the last class.  This training step can take a while (more than 5 minutes, depending on your computer).  To save some time, the instructors modified the code from last time to save the results of the training data to a file and saved this file in the class git repository.  You can used the saved data by setting ```rerun_training``` to `False` or, if you want, you can redo the training step by setting ```rerun_training``` to `True."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "rerun_training = True\n",
    "filename = 'full_face_model.p'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Train a SVM classification model  NOTE This can take ~ 5 minutes or more!!!!\n",
    "\n",
    "#make some temporary variables so you can change this easily\n",
    "tmp_vectors = train_vectors\n",
    "tmp_labels = train_labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "reading pickle file.\n",
      "Best estimator found by grid search:\n",
      "SVC(C=1000.0, class_weight='balanced', gamma=0.0001, kernel='linear')\n",
      "Runtime 0.00923776626586914\n"
     ]
    }
   ],
   "source": [
    "start = time.time()\n",
    "rerun_training = True\n",
    "if rerun_training:\n",
    "    \n",
    "    print(\"Fitting the classifier to the training set\")\n",
    "    param_grid = {'C': [1e3, 5e3, 1e4, 5e4, 1e5],\n",
    "                  'gamma': [0.0001, 0.0005, 0.001, 0.005, 0.01, 0.1], }\n",
    "    clf = GridSearchCV(SVC(kernel='linear', class_weight='balanced'), param_grid, n_jobs=-1)\n",
    "    clf = clf.fit(tmp_vectors, tmp_labels)\n",
    "    print(\"Best estimator found by grid search:\")\n",
    "    print(clf.best_estimator_)\n",
    "\n",
    "    #save the model to a file\n",
    "    pickle.dump(clf, open(filename, 'wb'))\n",
    "\n",
    "else:\n",
    "    #read the model from a file\n",
    "    print(\"reading pickle file.\")\n",
    "    clf = pickle.load(open(filename, 'rb')) #serialization\n",
    "    print(\"Best estimator found by grid search:\")\n",
    "    print(clf.best_estimator_)\n",
    "    \n",
    "end = time.time()\n",
    "print(\"Runtime\",end - start)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Question**: What library and functions are used to save and load a python object in this example?  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font size=8 color=\"#009600\">&#9998;</font> Do This - Erase the contents of this cell and replace it with your answer to the above question!  (double-click on this text to edit this cell, and hit shift+enter to save the text)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 3. Show the results of the classification on the testing dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Quantitative evaluation of the model quality on the test set\n",
    "\n",
    "#make some temporary variables so you can change this easily\n",
    "predict_vectors = test_vectors\n",
    "true_labels = test_labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predicting people's names on the test set\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.60      0.83      0.70        18\n",
      "           1       0.88      0.77      0.82        65\n",
      "           2       0.61      0.64      0.62        22\n",
      "           3       0.81      0.83      0.82       141\n",
      "           4       0.63      0.77      0.69        22\n",
      "           5       0.88      0.70      0.78        20\n",
      "           6       0.78      0.78      0.78         9\n",
      "           7       0.80      0.57      0.67        14\n",
      "           8       0.64      0.64      0.64        11\n",
      "           9       0.86      0.95      0.90        19\n",
      "          10       0.90      0.75      0.82        12\n",
      "          11       0.61      0.59      0.60        37\n",
      "\n",
      "    accuracy                           0.76       390\n",
      "   macro avg       0.75      0.73      0.74       390\n",
      "weighted avg       0.77      0.76      0.76       390\n",
      "\n",
      "[[ 15   0   0   2   0   0   0   1   0   0   0   0]\n",
      " [  3  50   2   6   0   0   1   0   0   0   1   2]\n",
      " [  3   0  14   4   0   0   0   0   0   0   0   1]\n",
      " [  2   5   3 117   4   1   1   0   1   2   0   5]\n",
      " [  0   0   1   2  17   0   0   0   0   0   0   2]\n",
      " [  1   0   0   2   1  14   0   0   0   1   0   1]\n",
      " [  0   1   0   1   0   0   7   0   0   0   0   0]\n",
      " [  1   0   2   2   0   0   0   8   1   0   0   0]\n",
      " [  0   0   0   1   0   0   0   0   7   0   0   3]\n",
      " [  0   0   0   0   1   0   0   0   0  18   0   0]\n",
      " [  0   0   1   0   0   1   0   1   0   0   9   0]\n",
      " [  0   1   0   8   4   0   0   0   2   0   0  22]]\n"
     ]
    }
   ],
   "source": [
    "print(\"Predicting people's names on the test set\")\n",
    "pred_labels = clf.predict(predict_vectors)\n",
    "\n",
    "print(classification_report(true_labels, pred_labels))\n",
    "print(confusion_matrix(true_labels, pred_labels, labels=range(n_classes)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x864 with 25 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_gallery(images, true_titles, pred_titles, h, w, n_row=5, n_col=5):\n",
    "    \"\"\"Helper function to plot a gallery of portraits\"\"\"\n",
    "    plt.figure(figsize=(1.8 * n_col, 2.4 * n_row))\n",
    "    plt.subplots_adjust(bottom=0, left=.01, right=.99, top=.90, hspace=.35)\n",
    "    for i in range(n_row * n_col):\n",
    "        plt.subplot(n_row, n_col, i + 1)\n",
    "        plt.imshow(images[i].reshape((h, w)), cmap=plt.cm.gray)\n",
    "        plt.title('Pred='+str(categories[pred_titles[i]]), size=9)\n",
    "        plt.xlabel('Actual='+str(categories[true_titles[i]]), size=9)\n",
    "        plt.xticks(())\n",
    "        plt.yticks(())\n",
    "\n",
    "plot_gallery(test_vectors, test_labels, pred_labels, h,w)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# 2. Review of pre-class assignment"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "\n",
    "# 3. Feature Selection using Principal Component Analysis\n",
    "\n",
    "Now we are going to go through the entire program above and use Principal Component Analysis (PCA) to select a much smaller set of better features.   The following code will reduce the original picture feature vector to a feature vector with 4 (n_components).  This is often called unsupervised feature extraction:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting the top 4 eigenfaces from 1170 faces\n"
     ]
    }
   ],
   "source": [
    "n_components = 4 # This is much less than the original n_features\n",
    "\n",
    "print(\"Extracting the top %d eigenfaces from %d faces\" % (n_components, train_vectors.shape[0]))\n",
    "\n",
    "#Set up the pca object with the number of compoents we want to find\n",
    "pca = PCA(n_components=n_components, whiten=True)\n",
    "\n",
    "#Fit the training data to the pca model.\n",
    "_ = pca.fit(train_vectors)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now have a ```pca``` object that has our data fit to it.  We can use this object to transform the original training vectors consisting of the entire image into new training vectors using the ```transform``` function:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training set changed from a size of:  (1170, 1850)  to:  (1170, 4)\n",
      "Testing set changed from a size of:  (390, 1850)  to:  (390, 4)\n"
     ]
    }
   ],
   "source": [
    "pca_train_vectors = pca.transform(train_vectors)\n",
    "pca_test_vectors = pca.transform(test_vectors)\n",
    "\n",
    "print(\"Training set changed from a size of: \", train_vectors.shape, ' to: ', pca_train_vectors.shape)\n",
    "print(\"Testing set changed from a size of: \", test_vectors.shape, ' to: ', pca_test_vectors.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "&#9989; **DO THIS:** We have now reduced the size of our feature vectors.  Modify the code in **Section 1** to substitute these two new reduced size vectors for the full size vectors.  This modification should replace both the training vectors and the testing vectors. Make sure you also consider the following:\n",
    "\n",
    "* Recalculate the clf model and do not load it from memory -- **this is important!**\n",
    "* You probably want to **change the filename so you do not overwrite the instructors file**\n",
    "* Make sure your changes ensure that **the tmp_vectors and predict_vectors reference the new smaller vectors**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Question**:  What precision can the SVM algorithm achieve with only the first 4 feature vectors?  And how long did this take to run?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font size=8 color=\"#009600\">&#9998;</font> Do This - Erase the contents of this cell and replace it with your answer to the above question!  (double-click on this text to edit this cell, and hit shift+enter to save the text)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Picking the \"best\" Feature Vectors\n",
    "If you remember from your pre-class assignment the PCA algorithm finds a transform of the data such that the first component contains the \"most\" information the second component contains the \"second most\" important information.  How much information each component contains is actually included in the ```pca``` object and can be expressed as a ratio from 0 (no information) to 1 (all information).  Lets plot these values below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x26d90417548>]"
      ]
     },
     "execution_count": 18,
     "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": [
    "#Lets plot the variance of the eigen values\n",
    "plt.plot(pca.explained_variance_ratio_, marker=\"o\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Another way to look at this is we can sum up the total ratios and see how much our new set of features represents the \"variance\" in the original data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "These 4 eigenvectors account for a total of 48 percent of the total variance in the original dataset\n"
     ]
    }
   ],
   "source": [
    "total_variance = np.sum(pca.explained_variance_ratio_)*100\n",
    "print(\"These %d eigenvectors account for a total of %d percent of the total variance in the original dataset\"\n",
    "      % (n_components, total_variance))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Question:** How many components would we need to represent 90% of the variance in our original data? (Hint: modify **Section 3** and change n_components to find a total_variance > 90)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font size=8 color=\"#009600\">&#9998;</font> Do This - Erase the contents of this cell and replace it with your answer to the above question!  (double-click on this text to edit this cell, and hit shift+enter to save the text)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "&#9989; **DO THIS:** generate a plot that generates a bar graph and compares the precision achieved of the following:\n",
    "* original - The original image as a complete feature vector.\n",
    "* 4 - The four \"best\" PCA components\n",
    "* 90% - The components representing the top 90% of the PCA variance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Put your code here\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "# 5. Eigenfaces (just for fun and knowledge)\n",
    "\n",
    "An eigenvector is a transform from the original image space into a new space along a single axis.  You can actually think of these vectors as a weighted sum of the components of an original image. So the length of the eigenvectors are the same size as the original data. Let's plot the gallery of the most significant eigenvectors:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "eigenfaces = pca.components_.reshape((n_components, h, w))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_gallery(images, titles, h, w, n_row=3, n_col=5):\n",
    "    \"\"\"Helper function to plot a gallery of portraits\"\"\"\n",
    "    plt.figure(figsize=(1.7 * n_col, 2.3 * n_row))\n",
    "    plt.subplots_adjust(bottom=0, left=.01, right=.99, top=.90, hspace=.35)\n",
    "    for i in range(n_row * n_col):\n",
    "        plt.subplot(n_row, n_col, i + 1)\n",
    "        plt.imshow(images[i].reshape((h, w)), cmap=plt.cm.gray)\n",
    "        plt.title(titles[i], size=12)\n",
    "        plt.xticks(())\n",
    "        plt.yticks(())\n",
    "        \n",
    "\n",
    "eigenface_titles = [\"eigenface %d\" % i for i in range(eigenfaces.shape[0])]\n",
    "plot_gallery(eigenfaces, eigenface_titles, h, w, n_row=1, n_col=4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These vectors are often called eigenfaces.  Each image gives you information about where the greatest amount of variation occurs in the images. Some of the eigenfaces tell you more about the lighting in the image or about the orientation of the face and less about the actual facial features, but others highlight where the most variation is from face to face."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-----\n",
    "### Congratulations, we're done!\n",
    "\n",
    "Now, you just need to submit this assignment by uploading it to the course <a href=\"https://d2l.msu.edu/\">Desire2Learn</a> web page for today's submission folder (Don't forget to add your names in the first cell).\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "&#169; Copyright Michigan State University Board of Trustees"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "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
}
