KNN_From_Scratch.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Only using numpy ,pandas and seaborn for visualization\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sepal length in cm</th>\n",
       "      <th>sepal width in cm</th>\n",
       "      <th>petal length in cm</th>\n",
       "      <th>petal width in cm</th>\n",
       "      <th>class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.1</td>\n",
       "      <td>3.5</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4.7</td>\n",
       "      <td>3.2</td>\n",
       "      <td>1.3</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.6</td>\n",
       "      <td>3.1</td>\n",
       "      <td>1.5</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>3.6</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sepal length in cm  sepal width in cm  petal length in cm  \\\n",
       "0                 5.1                3.5                 1.4   \n",
       "1                 4.9                3.0                 1.4   \n",
       "2                 4.7                3.2                 1.3   \n",
       "3                 4.6                3.1                 1.5   \n",
       "4                 5.0                3.6                 1.4   \n",
       "\n",
       "   petal width in cm        class  \n",
       "0                0.2  Iris-setosa  \n",
       "1                0.2  Iris-setosa  \n",
       "2                0.2  Iris-setosa  \n",
       "3                0.2  Iris-setosa  \n",
       "4                0.2  Iris-setosa  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Link to dataset : http://archive.ics.uci.edu/ml/datasets/iris\n",
    "\n",
    "cols=[\"sepal length in cm\",\"sepal width in cm\",\"petal length in cm\",\"petal width in cm\",\"class\"]\n",
    "df=pd.read_csv(\"http://archive.ics.uci.edu/ml/machine-learning-databases/iris/bezdekIris.data\",names=cols)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sepal length in cm</th>\n",
       "      <th>sepal width in cm</th>\n",
       "      <th>petal length in cm</th>\n",
       "      <th>petal width in cm</th>\n",
       "      <th>class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>6.7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.2</td>\n",
       "      <td>2.3</td>\n",
       "      <td>Iris-virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>5.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>1.6</td>\n",
       "      <td>0.6</td>\n",
       "      <td>Iris-setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>7.3</td>\n",
       "      <td>2.9</td>\n",
       "      <td>6.3</td>\n",
       "      <td>1.8</td>\n",
       "      <td>Iris-virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>7.7</td>\n",
       "      <td>3.8</td>\n",
       "      <td>6.7</td>\n",
       "      <td>2.2</td>\n",
       "      <td>Iris-virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>4.3</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.1</td>\n",
       "      <td>Iris-setosa</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     sepal length in cm  sepal width in cm  petal length in cm  \\\n",
       "145                 6.7                3.0                 5.2   \n",
       "43                  5.0                3.5                 1.6   \n",
       "107                 7.3                2.9                 6.3   \n",
       "117                 7.7                3.8                 6.7   \n",
       "13                  4.3                3.0                 1.1   \n",
       "\n",
       "     petal width in cm           class  \n",
       "145                2.3  Iris-virginica  \n",
       "43                 0.6     Iris-setosa  \n",
       "107                1.8  Iris-virginica  \n",
       "117                2.2  Iris-virginica  \n",
       "13                 0.1     Iris-setosa  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Give the dataset a good shuffle\n",
    "df=df.sample(frac=1)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Lets Split the data \n",
    "def train_test_split(X,y,test_frac=0.25):\n",
    "    X=X.values\n",
    "    y=y.values\n",
    "    X_train = X[:-int(len(X)*test_frac)]\n",
    "    X_test =  X[int(len(X)*test_frac):]\n",
    "    y_train = y[:-int(len(y)*test_frac)]\n",
    "    y_test =  y[int(len(y)*test_frac):]\n",
    "    return X_train,X_test,y_train,y_test\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initially Lets consider only 2 features for classification\n",
    "# Features -> sepal length and petal length\n",
    "X_train,X_test,y_train,y_test=train_test_split(df[[\"sepal length in cm\",\"petal length in cm\"]],df['class'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Visualizing the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f59c5808ac8>"
      ]
     },
     "execution_count": 27,
     "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": [
    "sns.scatterplot(x=X_train[:,0],y=X_train[:,1],hue=y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Lets define a distance function to calculate the distance between a feature point and test_sample\n",
    "t1,t2=X_train[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([6.7, 5.2, 3. ])"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t1 #Say our feature point"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([5. , 1.6, 3.5])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t2 #Say out target_sample point"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Calculating Euclidian Distance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "import math"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.981205847478877"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "math.sqrt( (t2[0]-t1[0])**2 + (t2[1]-t1[1])**2 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "#If we had 3 features we need to do sqrt( sum( (x2-x1)**2 , (y2-y1)**2 , (z2-z1)**2 ) )\n",
    "# Not so scalable if we need to keep rewriting the formula for each case ,\n",
    "# hence we use numpy linalg.norm() which does the same thing but faster and for any no of dimensions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_euclidian_distance(p1,p2):\n",
    "    \"\"\"\n",
    "    Input : \n",
    "        p1,p2 : Two points between which the distance is requried\n",
    "        \n",
    "    Returns : Euclidian distance. \n",
    "        i.e : The square-root of sum of squared difference of each dimension\n",
    "        \n",
    "    \"\"\"\n",
    "    return np.linalg.norm(p1-p2)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.012480529547776"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#We get the same result as before so we are good to proceed\n",
    "get_euclidian_distance(t1,t2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import Counter\n",
    "def KNN(data,test_case,k=5):\n",
    "    \"\"\"\n",
    "    data = data to train the classifier. expected a list of n elements where first (n-1) are features and last is class \n",
    "    test_case = features to predict a target label for\n",
    "    k = no of neighbours to use for calculating the distance\n",
    "    \n",
    "    Returns:\n",
    "        the class of the target label\n",
    "    \"\"\"\n",
    "    \n",
    "    #distance\n",
    "    distances=[]\n",
    "    labels=[]\n",
    "\n",
    "    for row in data:\n",
    "        #Each row of the data will contain the features and target class\n",
    "        features=row[:-1]\n",
    "        class_label=row[-1]\n",
    "        #First lets calculate the distance between the feature and test_case \n",
    "        dist=get_euclidian_distance(features,test_case)\n",
    "        #print(dist)\n",
    "        distances.append((dist,class_label))\n",
    "    \n",
    "    \n",
    "        for i in sorted(distances)[:k] :\n",
    "            labels.append(i[1]) \n",
    "\n",
    "    res=Counter(labels).most_common()[0][0]\n",
    "    \n",
    "    #res must only be a counter for the below to work\n",
    "    #acc=[]\n",
    "    #Calculating percentage on each class \n",
    "    #for r in res:\n",
    "    #    acc.append((r,res[r]/sum(res.values())))\n",
    "    \n",
    "    return res #acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [],
   "source": [
    "#data=df[['sepal length in cm','petal length in cm','class']].values\n",
    "data=np.hstack((X_train,y_train.reshape(len(X_train),1)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [],
   "source": [
    "res=KNN(data,[5,3],3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Iris-versicolor'"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "score=0\n",
    "for x,y in zip(X_test,y_test):\n",
    "    #print(x)\n",
    "    if KNN(data,x)==y:\n",
    "        score+=1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9557522123893806"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_acc=score/len(X_test)\n",
    "test_acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "# now we can use all the four feature columns and see how it performs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train,X_test,y_train,y_test=train_test_split(df.drop('class',axis=1),df['class'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [],
   "source": [
    "data=np.hstack((X_train,y_train.reshape(len(X_train),1)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9734513274336283"
      ]
     },
     "execution_count": 133,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score=0\n",
    "for x,y in zip(X_test,y_test):\n",
    "    #print(x)\n",
    "    if KNN(data,x)==y:\n",
    "        score+=1\n",
    "test_acc=score/len(X_test)\n",
    "test_acc"
   ]
  },
  {
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}