KNN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-15T06:24:22.028709Z",
     "start_time": "2019-03-15T06:24:21.942532Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_iris, load_diabetes\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from collections import Counter\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "% matplotlib inline\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-15T05:58:32.770251Z",
     "start_time": "2019-03-15T05:58:32.758294Z"
    }
   },
   "outputs": [],
   "source": [
    "iris = load_iris()\n",
    "X = iris.data\n",
    "Y = iris.target\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.33, random_state=42)\n",
    "sc = StandardScaler()\n",
    "X_train = sc.fit_transform(X_train)\n",
    "X_test = sc.transform(X_test)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-15T06:11:50.427221Z",
     "start_time": "2019-03-15T06:11:50.421444Z"
    }
   },
   "outputs": [],
   "source": [
    "class KnnClassifier:\n",
    "    def __init__(self, k,X_train, Y_train):\n",
    "        self.k = k\n",
    "        self.X_train = X_train\n",
    "        self.Y_train = Y_train\n",
    "        \n",
    "    def euclidean_distance(self,x_test):\n",
    "        nn_index = np.argsort(np.sqrt(np.sum(np.square(x_test - self.X_train), axis=1)))[0:self.k]\n",
    "        return nn_index\n",
    "\n",
    "    def majority_class(self,nn_index):\n",
    "        counts = Counter(self.Y_train[nn_index])\n",
    "        y_pred = counts.most_common(n=1)[0][0]\n",
    "\n",
    "        return y_pred\n",
    "\n",
    "    def fit(self, X_test):\n",
    "        y_preds = []\n",
    "        for x in X_test:\n",
    "            nn_index = self.euclidean_distance(x)\n",
    "            y_pred = self.majority_class(nn_index)\n",
    "            y_preds.append(y_pred)\n",
    "\n",
    "        return np.array(y_preds)\n",
    "    \n",
    "    def accuracy(self, y_pred, y_test):\n",
    "        return np.sum(y_pred == y_test)*1.0/len(y_pred)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Choose k with highest validation accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-15T06:11:37.448013Z",
     "start_time": "2019-03-15T06:11:36.962072Z"
    },
    "run_control": {
     "marked": true
    }
   },
   "outputs": [],
   "source": [
    "train_acc = []\n",
    "for i in range(2,100):\n",
    "    knn = KnnClassifier(i,X_train, Y_train)\n",
    "    y_preds = knn.fit(X_train)\n",
    "    train_acc.append(knn.accuracy(y_preds, Y_train))\n",
    "\n",
    "        \n",
    "    \n",
    "val_acc = []\n",
    "for i in range(2,100):\n",
    "    knn = KnnClassifier(i,X_train, Y_train)\n",
    "    y_preds = knn.fit(X_test)\n",
    "    val_acc.append(knn.accuracy(y_preds, Y_test))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-15T06:11:38.093501Z",
     "start_time": "2019-03-15T06:11:37.999994Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_range = list(range(2,100))\n",
    "\n",
    "train, = plt.plot(x_range, train_acc, c = 'g', label = 'train')\n",
    "val, = plt.plot(x_range, val_acc, c = 'b', label = 'validation')\n",
    "plt.vlines(x=3,ymin=0, ymax=1, colors='r', label = 'Highest Val Acc')\n",
    "plt.legend(handles=[train, val])\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- K = 3 gives highest validation accuracy for our dataset.\n",
    "- For smaller values of k have higher accuracy, larger value tends to generalize well on train data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-15T06:26:53.424386Z",
     "start_time": "2019-03-15T06:26:53.418768Z"
    }
   },
   "outputs": [],
   "source": [
    "class KnnRegressor:\n",
    "    def __init__(self, k,X_train, Y_train):\n",
    "        self.k = k\n",
    "        self.X_train = X_train\n",
    "        self.Y_train = Y_train\n",
    "        \n",
    "    def euclidean_distance(self,x_test):\n",
    "        nn_index = np.argsort(np.sqrt(np.sum(np.square(x_test - self.X_train), axis=1)))[0:self.k]\n",
    "        return nn_index\n",
    "\n",
    "    def mean_nn_score(self,nn_index):\n",
    "        y_pred = np.mean(self.Y_train[nn_index])\n",
    "\n",
    "        return y_pred\n",
    "\n",
    "    def fit(self, X_test):\n",
    "        y_preds = []\n",
    "        for x in X_test:\n",
    "            nn_index = self.euclidean_distance(x)\n",
    "            y_pred = self.mean_nn_score(nn_index)\n",
    "            y_preds.append(y_pred)\n",
    "\n",
    "        return np.array(y_preds)\n",
    "    \n",
    "    def rmse(self, y_pred, y_test):\n",
    "        return round(np.sqrt(np.sum((y_pred - y_test)**2)),4)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-15T06:26:53.769342Z",
     "start_time": "2019-03-15T06:26:53.763164Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "525688.1111111111"
      ]
     },
     "execution_count": 200,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sum((y_preds-Y_test)**2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-15T06:28:49.705432Z",
     "start_time": "2019-03-15T06:28:49.686112Z"
    },
    "run_control": {
     "marked": true
    }
   },
   "outputs": [],
   "source": [
    "diabetes = load_diabetes()\n",
    "X = diabetes.data\n",
    "Y = diabetes.target\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.33, random_state=42)\n",
    "sc = StandardScaler()\n",
    "X_train = sc.fit_transform(X_train)\n",
    "X_test = sc.transform(X_test)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-15T06:28:50.046252Z",
     "start_time": "2019-03-15T06:28:50.040979Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((296, 10), (296,), (146, 10), (146,))"
      ]
     },
     "execution_count": 208,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, Y_train.shape, X_test.shape,Y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-15T06:28:51.808994Z",
     "start_time": "2019-03-15T06:28:51.797646Z"
    }
   },
   "outputs": [],
   "source": [
    "knn = KnnRegressor(3,X_train, Y_train)\n",
    "y_preds = knn.fit(X_test)\n",
    "val_rmse = knn.rmse(y_preds, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-03-15T06:31:04.104820Z",
     "start_time": "2019-03-15T06:31:02.228861Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_rmse = []\n",
    "for i in range(2,100):\n",
    "    knn = KnnRegressor(i,X_train, Y_train)\n",
    "    y_preds = knn.fit(X_train)\n",
    "    train_rmse.append(knn.rmse(y_preds, Y_train))\n",
    "\n",
    "        \n",
    "    \n",
    "val_rmse = []\n",
    "for i in range(2,100):\n",
    "    knn = KnnRegressor(i,X_train, Y_train)\n",
    "    y_preds = knn.fit(X_test)\n",
    "    val_rmse.append(knn.rmse(y_preds, Y_test))\n",
    "\n",
    "\n",
    "x_range = list(range(2,100))\n",
    "\n",
    "train, = plt.plot(x_range, train_rmse, c = 'g', label = 'train')\n",
    "val, = plt.plot(x_range, val_rmse, c = 'b', label = 'validation')\n",
    "plt.vlines(x=x_range[np.argsort(np.array(val_rmse))[0]],ymin=0, ymax=2000, colors='r', label = 'Highest Val Acc')\n",
    "plt.legend(handles=[train, val])\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.2"
  },
  "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
}