classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prepare the data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Jupyter theme settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:27:36.447195Z",
     "start_time": "2019-04-25T18:27:35.951211Z"
    }
   },
   "outputs": [],
   "source": [
    "# import jtplot module in notebook\n",
    "from jupyterthemes import jtplot\n",
    "\n",
    "# choose which theme to inherit plotting style from\n",
    "# onedork | grade3 | oceans16 | chesterish | monokai | solarizedl | solarizedd\n",
    "jtplot.style(theme='grade3', context='paper', fscale=1.5, spines=True, ticks=False)\n",
    "\n",
    "#jtplot.style(context='talk', fscale=1.4, spines=False, gridlines='--')\n",
    "#jtplot.style(ticks=True, grid=False, figsize=(6, 4.5))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Standard imports and functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:27:36.453656Z",
     "start_time": "2019-04-25T18:27:36.448497Z"
    }
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import os\n",
    "\n",
    "# to make output stable across runs\n",
    "np.random.seed(42)\n",
    "\n",
    "%matplotlib inline\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "#mpl.rc('axes', lablesize=14)\n",
    "mpl.rc('xtick', labelsize=12)\n",
    "mpl.rc('ytick', labelsize=12)\n",
    "\n",
    "# location to save figures\n",
    "PROJECT_ROOT_DIR = \".\"\n",
    "CHAPTER_ID = \"classification\"\n",
    "\n",
    "def save_fig(fig_id, tight_layout=True):\n",
    "    path = os.path.join(PROJECT_ROOT_DIR, \"images/\"+ CHAPTER_ID, fig_id + \".png\")\n",
    "    print(\"saving figure\", fig_id)\n",
    "    if tight_layout:\n",
    "        plt.tight_layout()\n",
    "    plt.savefig(path, format=\"png\", dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:27:36.463805Z",
     "start_time": "2019-04-25T18:27:36.454753Z"
    }
   },
   "outputs": [],
   "source": [
    "# Ignore useless warnings (see SciPy issue #5998)\n",
    "import warnings\n",
    "warnings.filterwarnings(action=\"ignore\", message=\"^internal gelsd\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:27:36.470554Z",
     "start_time": "2019-04-25T18:27:36.465052Z"
    }
   },
   "outputs": [],
   "source": [
    "def sort_by_target(mnist):\n",
    "    reorder_train = np.array(sorted([(target, i) for i, target in enumerate(mnist.target[:60000])]))[:, 1]\n",
    "    reorder_test = np.array(sorted([(target, i) for i, target in enumerate(mnist.target[60000:])]))[:, 1]\n",
    "    mnist.data[:60000] = mnist.data[reorder_train]\n",
    "    mnist.target[:60000] = mnist.target[reorder_train]\n",
    "    mnist.data[60000:] = mnist.data[reorder_test + 60000]\n",
    "    mnist.target[60000:] = mnist.target[reorder_test + 60000]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:27:36.521246Z",
     "start_time": "2019-04-25T18:27:36.471949Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.20.0'"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import sklearn\n",
    "sklearn.__version__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Download MNIST data from sklearn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:06.729608Z",
     "start_time": "2019-04-25T18:27:36.523233Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([[0., 0., 0., ..., 0., 0., 0.],\n",
       "        [0., 0., 0., ..., 0., 0., 0.],\n",
       "        [0., 0., 0., ..., 0., 0., 0.],\n",
       "        ...,\n",
       "        [0., 0., 0., ..., 0., 0., 0.],\n",
       "        [0., 0., 0., ..., 0., 0., 0.],\n",
       "        [0., 0., 0., ..., 0., 0., 0.]]),\n",
       " array([0, 0, 0, ..., 9, 9, 9], dtype=int8))"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "try:\n",
    "    from sklearn.datasets import fetch_openml\n",
    "    mnist = fetch_openml('mnist_784', version=1, cache=False)\n",
    "    mnist.target = mnist.target.astype(np.int8) # fetch_openml() returns targets as strings\n",
    "    sort_by_target(mnist) # fetch_openml() returns an unsorted dataset\n",
    "except ImportError:\n",
    "    from sklearn.datasets import fetch_mldata\n",
    "    mnist = fetch_mldata('MNIST original')\n",
    "mnist[\"data\"], mnist[\"target\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:06.733204Z",
     "start_time": "2019-04-25T18:28:06.730883Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(70000, 784)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mnist.data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:06.741909Z",
     "start_time": "2019-04-25T18:28:06.734497Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(70000, 784)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X, y = mnist[\"data\"], mnist[\"target\"]\n",
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:06.748510Z",
     "start_time": "2019-04-25T18:28:06.743378Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(70000,)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Each image is ```28*28``` pixels and each feature represents one pixel's intensity, from 0(white) to 255(black)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:06.756103Z",
     "start_time": "2019-04-25T18:28:06.749636Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "784"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "28*28"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "imshow: \n",
    "displays an image, e.e. data on a 2D regular raster.   \n",
    "cmap is Colormap.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:07.063597Z",
     "start_time": "2019-04-25T18:28:06.759009Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving figure some_digit_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "some_digit = X[36000]\n",
    "some_digit_image = some_digit.reshape(28, 28)\n",
    "plt.imshow(some_digit_image, cmap=mpl.cm.binary, interpolation=\"nearest\")\n",
    "plt.axis(\"off\")\n",
    "\n",
    "save_fig(\"some_digit_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "function to plot the digits."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:07.073025Z",
     "start_time": "2019-04-25T18:28:07.066103Z"
    }
   },
   "outputs": [],
   "source": [
    "def plot_digit(data):\n",
    "    image = data.reshape(28, 28)\n",
    "    plt.imshow(image, cmap=mpl.cm.binary, interpolation=\"nearest\")\n",
    "    plt.axis(\"off\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:07.083113Z",
     "start_time": "2019-04-25T18:28:07.075369Z"
    }
   },
   "outputs": [],
   "source": [
    "def plot_digits(instances, images_per_row=10, **options):\n",
    "    size = 28\n",
    "    images_per_row = min(len(instances), images_per_row)\n",
    "    images = [instance.reshape(size,size) for instance in instances]\n",
    "    n_rows = (len(instances) - 1) // images_per_row + 1\n",
    "    row_images = []\n",
    "    n_empty = n_rows * images_per_row - len(instances)\n",
    "    images.append(np.zeros((size, size * n_empty)))\n",
    "    for row in range(n_rows):\n",
    "        rimages = images[row * images_per_row : (row + 1) * images_per_row]\n",
    "        row_images.append(np.concatenate(rimages, axis=1))\n",
    "    image = np.concatenate(row_images, axis=0)\n",
    "    plt.imshow(image, cmap = mpl.cm.binary, **options)\n",
    "    plt.axis(\"off\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:07.988089Z",
     "start_time": "2019-04-25T18:28:07.085965Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving figure more_digits_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x648 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(9,9))\n",
    "example_images = np.r_[X[:12000:600], X[13000:30600:600], X[30600:60000:590]]\n",
    "plot_digits(example_images, images_per_row=10)\n",
    "save_fig(\"more_digits_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:07.994397Z",
     "start_time": "2019-04-25T18:28:07.990746Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y[36000]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Out of 70000 rows, we split fiest 60000 for training data and the rest for test."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:08.003764Z",
     "start_time": "2019-04-25T18:28:07.996066Z"
    }
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = X[:60000], X[60000:], y[:60000], y[60000:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets shuffle training set; that will guarantee that all cross-validation folds will be similar.    \n",
    "np.random.permutation : randomly permuate a sequence. Ex:   \n",
    "```\n",
    ">>> np.random.permutation([1, 4, 9, 12, 15])\n",
    "array([15,  1,  9,  4, 12])\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:09.162872Z",
     "start_time": "2019-04-25T18:28:08.006061Z"
    }
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "\n",
    "shuffle_index = np.random.permutation(60000)\n",
    "X_train, y_train = X_train[shuffle_index], y_train[shuffle_index]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Binary Classifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Training a classifier for predicting digit 5."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:09.168318Z",
     "start_time": "2019-04-25T18:28:09.164849Z"
    }
   },
   "outputs": [],
   "source": [
    "y_train_5 = (y_train == 5)\n",
    "y_test_5 = (y_test == 5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "SGDClassifier:    \n",
    "\n",
    "Linear classifiers (SVM, logistic regression, a.o.) with SGD training. This estimator implements regularized linear models with stochastic gradient descent (SGD) learning: the gradient of the loss is estimated each sample at a time and the model is updated along the way with a decreasing strength schedule (aka learning rate). This implementation works with data represented as dense or sparse arrays of floating point values for the features. The model it fits can be controlled with the loss parameter; by default, it fits a linear support vector machine (SVM)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:09.633510Z",
     "start_time": "2019-04-25T18:28:09.170092Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SGDClassifier(alpha=0.0001, average=False, class_weight=None,\n",
       "       early_stopping=False, epsilon=0.1, eta0=0.0, fit_intercept=True,\n",
       "       l1_ratio=0.15, learning_rate='optimal', loss='hinge', max_iter=5,\n",
       "       n_iter=None, n_iter_no_change=5, n_jobs=None, penalty='l2',\n",
       "       power_t=0.5, random_state=42, shuffle=True, tol=-inf,\n",
       "       validation_fraction=0.1, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import SGDClassifier\n",
    "\n",
    "sgd_clf = SGDClassifier(max_iter=5, tol=-np.infty, random_state=42)\n",
    "sgd_clf.fit(X_train, y_train_5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:09.639827Z",
     "start_time": "2019-04-25T18:28:09.635823Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ True])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sgd_clf.predict([some_digit])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:11.032838Z",
     "start_time": "2019-04-25T18:28:09.641530Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.9502 , 0.96565, 0.96495])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "cross_val_score(sgd_clf, X_train, y_train_5, cv=3, scoring=\"accuracy\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Writing function to calculate cross validation. This is just to show that instead of using cross val score we could write our own method and modify accordingly."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:12.522226Z",
     "start_time": "2019-04-25T18:28:11.034157Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9502\n",
      "0.96565\n",
      "0.96495\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.base import clone\n",
    "\n",
    "skfolds = StratifiedKFold(n_splits=3, random_state=42)\n",
    "\n",
    "for train_index, test_index in skfolds.split(X_train, y_train_5):\n",
    "    clone_clf = clone(sgd_clf)\n",
    "    X_train_folds = X_train[train_index]\n",
    "    y_train_folds = (y_train_5[train_index])\n",
    "    X_test_fold = X_train[test_index]\n",
    "    y_test_fold = (y_train_5[test_index])\n",
    "\n",
    "    clone_clf.fit(X_train_folds, y_train_folds)\n",
    "    y_pred = clone_clf.predict(X_test_fold)\n",
    "    n_correct = sum(y_pred == y_test_fold)\n",
    "    print(n_correct / len(y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Simple classifier to predict if it's not a 5."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:12.528556Z",
     "start_time": "2019-04-25T18:28:12.525071Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.base import BaseEstimator\n",
    "\n",
    "class Never5Classifier(BaseEstimator):\n",
    "    def fit(self, X, y=None):\n",
    "        pass\n",
    "    def predict(self, X):\n",
    "        return np.zeros((len(X), 1), dtype=bool)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:13.005705Z",
     "start_time": "2019-04-25T18:28:12.530354Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.909  , 0.90715, 0.9128 ])"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "never_5_clf = Never5Classifier()\n",
    "cross_val_score(never_5_clf, X_train, y_train_5, cv=3, scoring=\"accuracy\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.447006Z",
     "start_time": "2019-04-25T18:28:13.008694Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import cross_val_predict\n",
    "\n",
    "y_train_pred = cross_val_predict(sgd_clf, X_train, y_train_5, cv=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.503748Z",
     "start_time": "2019-04-25T18:28:14.449187Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[53272,  1307],\n",
       "       [ 1077,  4344]])"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "confusion_matrix(y_train_5, y_train_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.507721Z",
     "start_time": "2019-04-25T18:28:14.505436Z"
    }
   },
   "outputs": [],
   "source": [
    "y_train_perfect_predictions = y_train_5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.543198Z",
     "start_time": "2019-04-25T18:28:14.509085Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[54579,     0],\n",
       "       [    0,  5421]])"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "confusion_matrix(y_train_5, y_train_perfect_predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.557033Z",
     "start_time": "2019-04-25T18:28:14.544951Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7687135020350381"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import precision_score, recall_score\n",
    "\n",
    "precision_score(y_train_5, y_train_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "precision = TP/(TP + FP)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.562454Z",
     "start_time": "2019-04-25T18:28:14.559433Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7687135020350381"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "4344/(4344+1307)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.578237Z",
     "start_time": "2019-04-25T18:28:14.563717Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.801328168234643"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "recall_score(y_train_5, y_train_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "recall = TP/(TP + FN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.583011Z",
     "start_time": "2019-04-25T18:28:14.579498Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.801328168234643"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "4344/(4344+1077)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.596581Z",
     "start_time": "2019-04-25T18:28:14.584365Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7846820809248555"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import f1_score\n",
    "f1_score(y_train_5, y_train_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "f1score = TP/(TP + (FN+FP)/2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.601377Z",
     "start_time": "2019-04-25T18:28:14.598185Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7846820809248555"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "4344 / (4344 + (1077 + 1307)/2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Scikit-Learn does not let you set the threshold directly, but it does give you access to\n",
    "the decision scores that it uses to make predictions. Instead of calling the classifier’s\n",
    "predict() method, you can call its decision_function() method, which returns a\n",
    "score for each instance, and then make predictions based on those scores using any\n",
    "threshold you want:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.609072Z",
     "start_time": "2019-04-25T18:28:14.602956Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([161855.74572176])"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_scores = sgd_clf.decision_function([some_digit])\n",
    "y_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.615475Z",
     "start_time": "2019-04-25T18:28:14.610300Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ True])"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "threshold = 0\n",
    "y_some_digit_pred = (y_scores > threshold)\n",
    "y_some_digit_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:14.622290Z",
     "start_time": "2019-04-25T18:28:14.616684Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([False])"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "threshold = 200000\n",
    "y_some_digit_pred = (y_scores > threshold)\n",
    "y_some_digit_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**cross_val_predict:   **\n",
    "\n",
    "Generate cross-validated estimates for each input data point"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:16.043323Z",
     "start_time": "2019-04-25T18:28:14.623387Z"
    }
   },
   "outputs": [],
   "source": [
    "y_scores = cross_val_predict(sgd_clf, X_train, y_train_5, cv=3, method=\"decision_function\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:16.047309Z",
     "start_time": "2019-04-25T18:28:16.044899Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000,)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_scores.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:16.058165Z",
     "start_time": "2019-04-25T18:28:16.048713Z"
    }
   },
   "outputs": [],
   "source": [
    "# hack to work around issue #9589 in Scikit-Learn 0.19.0\n",
    "# if y_scores.ndim == 2:\n",
    "#     y_scores = y_scores[:, 1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**precision_recall_curve:**   \n",
    "Compute precision-recall pairs for different probability thresholds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:16.077644Z",
     "start_time": "2019-04-25T18:28:16.060868Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import precision_recall_curve\n",
    "\n",
    "precisions, recalls, thresholds = precision_recall_curve(y_train_5, y_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:16.402079Z",
     "start_time": "2019-04-25T18:28:16.079212Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving figure precision_recall_vs_threshold_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_precision_recall_vs_threshold(precisions, recalls, thresholds):\n",
    "    plt.plot(thresholds, precisions[:-1], \"b--\", label=\"Precision\", linewidth=2)\n",
    "    plt.plot(thresholds, recalls[:-1], \"g-\", label=\"Recall\", linewidth=2)\n",
    "    plt.xlabel(\"Threshold\", fontsize=16)\n",
    "    plt.legend(loc=\"upper left\", fontsize=16)\n",
    "    plt.ylim([0, 1])\n",
    "\n",
    "plt.figure(figsize=(8, 4))\n",
    "plot_precision_recall_vs_threshold(precisions, recalls, thresholds)\n",
    "plt.xlim([-700000, 700000])\n",
    "save_fig(\"precision_recall_vs_threshold_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:16.408197Z",
     "start_time": "2019-04-25T18:28:16.404191Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(y_train_pred == (y_scores > 0)).all()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "aim for 90% precision and lets say we can achieve that at threshold 70000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:16.418686Z",
     "start_time": "2019-04-25T18:28:16.409912Z"
    }
   },
   "outputs": [],
   "source": [
    "y_train_pred_90 = (y_scores > 70000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:16.433519Z",
     "start_time": "2019-04-25T18:28:16.420768Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8659205116491548"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "precision_score(y_train_5, y_train_pred_90)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:16.446807Z",
     "start_time": "2019-04-25T18:28:16.435674Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6993174691016417"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "recall_score(y_train_5, y_train_pred_90)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "from above we can say that we got a 90% precision classifier but on the trade off of a low recall."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:16.799596Z",
     "start_time": "2019-04-25T18:28:16.448413Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving figure precision_vs_recall_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_precision_vs_recall(precisions, recalls):\n",
    "    plt.plot(recalls, precisions, \"b-\", linewidth=2)\n",
    "    plt.xlabel(\"Recall\", fontsize=16)\n",
    "    plt.ylabel(\"Precision\", fontsize=16)\n",
    "    plt.axis([0, 1, 0, 1])\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "plot_precision_vs_recall(precisions, recalls)\n",
    "save_fig(\"precision_vs_recall_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:16.815544Z",
     "start_time": "2019-04-25T18:28:16.802021Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import roc_curve\n",
    "\n",
    "fpr, tpr, thresholds = roc_curve(y_train_5, y_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:17.150835Z",
     "start_time": "2019-04-25T18:28:16.817806Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving figure roc_curve_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_roc_curve(fpr, tpr, label=None):\n",
    "    plt.plot(fpr, tpr, linewidth=2, label=label)\n",
    "    plt.plot([0, 1], [0, 1], 'k--')\n",
    "    plt.axis([0, 1, 0, 1])\n",
    "    plt.xlabel('False Positive Rate', fontsize=16)\n",
    "    plt.ylabel('True Positive Rate', fontsize=16)\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "plot_roc_curve(fpr, tpr)\n",
    "save_fig(\"roc_curve_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:17.173442Z",
     "start_time": "2019-04-25T18:28:17.153161Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9624496555967156"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "roc_auc_score(y_train_5, y_scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note: we set n_estimators=10 to avoid a warning about the fact that its default value will be set to 100 in Scikit-Learn 0.22."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:25.124730Z",
     "start_time": "2019-04-25T18:28:17.175142Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "forest_clf = RandomForestClassifier(n_estimators=10, random_state=42)\n",
    "y_probas_forest = cross_val_predict(forest_clf, X_train, y_train_5, cv=3, method=\"predict_proba\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:25.134238Z",
     "start_time": "2019-04-25T18:28:25.127037Z"
    }
   },
   "outputs": [],
   "source": [
    "y_scores_forest = y_probas_forest[:, 1] # score = proba of positive class\n",
    "fpr_forest, tpr_forest, thresholds_forest = roc_curve(y_train_5,y_scores_forest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:25.494840Z",
     "start_time": "2019-04-25T18:28:25.135557Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving figure roc_curve_comparison_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8, 6))\n",
    "plt.plot(fpr, tpr, \"b:\", linewidth=2, label=\"SGD\")\n",
    "plot_roc_curve(fpr_forest, tpr_forest, \"Random Forest\")\n",
    "plt.legend(loc=\"lower right\", fontsize=16)\n",
    "save_fig(\"roc_curve_comparison_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:25.510720Z",
     "start_time": "2019-04-25T18:28:25.496839Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9931243366003829"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "roc_auc_score(y_train_5, y_scores_forest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:33.333006Z",
     "start_time": "2019-04-25T18:28:25.512628Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9852973447443494"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train_pred_forest = cross_val_predict(forest_clf, X_train, y_train_5, cv=3)\n",
    "precision_score(y_train_5, y_train_pred_forest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:33.350382Z",
     "start_time": "2019-04-25T18:28:33.334890Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8282604685482383"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "recall_score(y_train_5, y_train_pred_forest)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multiclass classification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Multiclass classification** means a classification task with more than two classes; e.g., classify a set of images of fruits which may be oranges, apples, or pears. Multiclass classification makes the assumption that each sample is assigned to one and only one label: a fruit can be either an apple or a pear but not both at the same time.     \n",
    "\n",
    "**Multilabel classification** assigns to each sample a set of target labels. This can be thought as predicting properties of a data-point that are not mutually exclusive, such as topics that are relevant for a document. A text might be about any of religion, politics, finance or education at the same time or none of these.    \n",
    "\n",
    "**Multioutput regression** assigns each sample a set of target values. This can be thought of as predicting several properties for each data-point, such as wind direction and magnitude at a certain location.     \n",
    "\n",
    "**Multioutput-multiclass classification** and multi-task classification means that a single estimator has to handle several joint classification tasks. This is both a generalization of the multi-label classification task, which only considers binary classification, as well as a generalization of the multi-class classification task. The output format is a 2d numpy array or sparse matrix.   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:37.228799Z",
     "start_time": "2019-04-25T18:28:33.352131Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([5], dtype=int8)"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sgd_clf.fit(X_train, y_train)\n",
    "sgd_clf.predict([some_digit])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:37.235277Z",
     "start_time": "2019-04-25T18:28:37.230224Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-311402.62954431, -363517.28355739, -446449.5306454 ,\n",
       "        -183226.61023518, -414337.15339485,  161855.74572176,\n",
       "        -452576.39616343, -471957.14962573, -518542.33997148,\n",
       "        -536774.63961222]])"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "some_digit_scores = sgd_clf.decision_function([some_digit])\n",
    "some_digit_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:37.247340Z",
     "start_time": "2019-04-25T18:28:37.237027Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argmax(some_digit_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:37.254602Z",
     "start_time": "2019-04-25T18:28:37.249610Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9], dtype=int8)"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sgd_clf.classes_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:37.262683Z",
     "start_time": "2019-04-25T18:28:37.256233Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sgd_clf.classes_[5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**OneVsOneClassifier:**    \n",
    "\n",
    "One-vs-one multiclass strategy\n",
    "\n",
    "This strategy consists in fitting one classifier per class pair. At prediction time, the class which received the most votes is selected. Since it requires to fit n_classes * (n_classes - 1) / 2 classifiers, this method is usually slower than one-vs-the-rest, due to its O(n_classes^2) complexity. However, this method may be advantageous for algorithms such as kernel algorithms which don’t scale well with n_samples. This is because each individual learning problem only involves a small subset of the data whereas, with one-vs-the-rest, the complete dataset is used n_classes times."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:42.406938Z",
     "start_time": "2019-04-25T18:28:37.264434Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([5], dtype=int8)"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.multiclass import OneVsOneClassifier\n",
    "\n",
    "ovo_clf = OneVsOneClassifier(SGDClassifier(max_iter=5, tol=-np.infty, random_state=42))\n",
    "ovo_clf.fit(X_train, y_train)\n",
    "ovo_clf.predict([some_digit])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:42.412578Z",
     "start_time": "2019-04-25T18:28:42.408948Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "45"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(ovo_clf.estimators_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:47.156438Z",
     "start_time": "2019-04-25T18:28:42.414033Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([5], dtype=int8)"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forest_clf.fit(X_train, y_train)\n",
    "forest_clf.predict([some_digit])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:47.164323Z",
     "start_time": "2019-04-25T18:28:47.158233Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.1, 0. , 0. , 0.1, 0. , 0.8, 0. , 0. , 0. , 0. ]])"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forest_clf.predict_proba([some_digit])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:28:55.144441Z",
     "start_time": "2019-04-25T18:28:47.167126Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.84063187, 0.84899245, 0.86652998])"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_val_score(sgd_clf, X_train, y_train, cv=3, scoring=\"accuracy\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:04.429085Z",
     "start_time": "2019-04-25T18:28:55.145984Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.91011798, 0.90874544, 0.906636  ])"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "scaler = StandardScaler()\n",
    "X_train_scaled = scaler.fit_transform(X_train.astype(np.float64))\n",
    "cross_val_score(sgd_clf, X_train_scaled, y_train, cv=3, scoring=\"accuracy\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:12.427256Z",
     "start_time": "2019-04-25T18:29:04.430817Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[5725,    3,   24,    9,   10,   49,   50,   10,   39,    4],\n",
       "       [   2, 6493,   43,   25,    7,   40,    5,   10,  109,    8],\n",
       "       [  51,   41, 5321,  104,   89,   26,   87,   60,  166,   13],\n",
       "       [  47,   46,  141, 5342,    1,  231,   40,   50,  141,   92],\n",
       "       [  19,   29,   41,   10, 5366,    9,   56,   37,   86,  189],\n",
       "       [  73,   45,   36,  193,   64, 4582,  111,   30,  193,   94],\n",
       "       [  29,   34,   44,    2,   42,   85, 5627,   10,   45,    0],\n",
       "       [  25,   24,   74,   32,   54,   12,    6, 5787,   15,  236],\n",
       "       [  52,  161,   73,  156,   10,  163,   61,   25, 5027,  123],\n",
       "       [  43,   35,   26,   92,  178,   28,    2,  223,   82, 5240]])"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train_pred = cross_val_predict(sgd_clf, X_train_scaled, y_train, cv=3)\n",
    "conf_mx = confusion_matrix(y_train, y_train_pred)\n",
    "conf_mx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:12.431334Z",
     "start_time": "2019-04-25T18:29:12.428743Z"
    }
   },
   "outputs": [],
   "source": [
    "def plot_confusion_matrix(matrix):\n",
    "    fig = plt.figure(figsize=(8,8))\n",
    "    ax = fig.add_subplot(111)\n",
    "    cax = ax.matshow(matrix)\n",
    "    fig.colorbar(cax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:12.648516Z",
     "start_time": "2019-04-25T18:29:12.432785Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving figure confusion_matrix_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.matshow(conf_mx, cmap=plt.cm.gray)\n",
    "save_fig(\"confusion_matrix_plot\", tight_layout=False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To compare error rates: need to divide each value in the confusion matrix by the number of images in the corresponding class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:12.652995Z",
     "start_time": "2019-04-25T18:29:12.650297Z"
    }
   },
   "outputs": [],
   "source": [
    "row_sums = conf_mx.sum(axis=1, keepdims=True)\n",
    "norm_conf_mx = conf_mx / row_sums"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:12.894187Z",
     "start_time": "2019-04-25T18:29:12.654871Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving figure confusion_matrix_errors_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "np.fill_diagonal(norm_conf_mx, 0)\n",
    "plt.matshow(norm_conf_mx, cmap=plt.cm.gray)\n",
    "save_fig(\"confusion_matrix_errors_plot\", tight_layout=False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:13.700011Z",
     "start_time": "2019-04-25T18:29:12.895556Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving figure error_analysis_digits_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cl_a, cl_b = 3, 5\n",
    "X_aa = X_train[(y_train == cl_a) & (y_train_pred == cl_a)]\n",
    "X_ab = X_train[(y_train == cl_a) & (y_train_pred == cl_b)]\n",
    "X_ba = X_train[(y_train == cl_b) & (y_train_pred == cl_a)]\n",
    "X_bb = X_train[(y_train == cl_b) & (y_train_pred == cl_b)]\n",
    "\n",
    "plt.figure(figsize=(8,8))\n",
    "plt.subplot(221); plot_digits(X_aa[:25], images_per_row=5)\n",
    "plt.subplot(222); plot_digits(X_ab[:25], images_per_row=5)\n",
    "plt.subplot(223); plot_digits(X_ba[:25], images_per_row=5)\n",
    "plt.subplot(224); plot_digits(X_bb[:25], images_per_row=5)\n",
    "save_fig(\"error_analysis_digits_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multilabel classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:47.161273Z",
     "start_time": "2019-04-25T18:29:13.701777Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "           metric_params=None, n_jobs=None, n_neighbors=5, p=2,\n",
       "           weights='uniform')"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "y_train_large = (y_train >= 7)\n",
    "y_train_odd = (y_train % 2 == 1)\n",
    "y_multilabel = np.c_[y_train_large, y_train_odd]\n",
    "\n",
    "knn_clf = KNeighborsClassifier()\n",
    "knn_clf.fit(X_train, y_multilabel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:47.255688Z",
     "start_time": "2019-04-25T18:29:47.163159Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[False,  True]])"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn_clf.predict([some_digit])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:47.260136Z",
     "start_time": "2019-04-25T18:29:47.257656Z"
    }
   },
   "outputs": [],
   "source": [
    "# y_train_knn_pred = cross_val_predict(knn_clf, X_train, y_multilabel, cv=3, n_jobs=-1)\n",
    "# f1_score(y_multilabel, y_train_knn_pred, average=\"macro\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multioutput classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:48.119611Z",
     "start_time": "2019-04-25T18:29:47.262918Z"
    }
   },
   "outputs": [],
   "source": [
    "noise = np.random.randint(0, 100, (len(X_train), 784))\n",
    "X_train_mod = X_train + noise\n",
    "noise = np.random.randint(0, 100, (len(X_test), 784))\n",
    "X_test_mod = X_test + noise\n",
    "y_train_mod = X_train\n",
    "y_test_mod = X_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:29:48.386589Z",
     "start_time": "2019-04-25T18:29:48.121842Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving figure noisy_digit_example_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "some_index = 5500\n",
    "plt.subplot(121); plot_digit(X_test_mod[some_index])\n",
    "plt.subplot(122); plot_digit(y_test_mod[some_index])\n",
    "save_fig(\"noisy_digit_example_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-04-25T18:30:01.132302Z",
     "start_time": "2019-04-25T18:29:48.389269Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving figure cleaned_digit_example_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "knn_clf.fit(X_train_mod, y_train_mod)\n",
    "clean_digit = knn_clf.predict([X_test_mod[some_index]])\n",
    "plot_digit(clean_digit)\n",
    "save_fig(\"cleaned_digit_example_plot\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---------------------------------------------------END-------------------------------------------------------------"
   ]
  }
 ],
 "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": "3.5.5"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}