KNN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.datasets import make_blobs\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = np.load('./mnist_train_small.npy') # we have take mnist data set."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have dimensions of approx 20K rows and 785 columns. Out of 785 columns, 1 column is our y i.e Output and remaining 784 i.e. image of 28*28 pixels is our examples or input.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[5, 0, 0, ..., 0, 0, 0],\n",
       "       [7, 0, 0, ..., 0, 0, 0],\n",
       "       [9, 0, 0, ..., 0, 0, 0],\n",
       "       ...,\n",
       "       [2, 0, 0, ..., 0, 0, 0],\n",
       "       [9, 0, 0, ..., 0, 0, 0],\n",
       "       [5, 0, 0, ..., 0, 0, 0]], dtype=uint8)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(19999, 785)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data[:,1:] # All the rows and column from 1 to 785. \n",
    "y = data[:,0] # All the rows and column with index 0."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "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]], dtype=uint8)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([5, 7, 9, ..., 2, 9, 5], dtype=uint8)"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((19999, 784), (19999,))"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape, y.shape # for each x examples we will have y answers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x25c1cb76580>"
      ]
     },
     "execution_count": 17,
     "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": [
    "plt.imshow(X[0].reshape(28,28), cmap ='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "...     X, Y, test_size=0.33, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((6600, 784), (6600,))"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((13399, 784), (13399,))"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier # this is inbuilt class.\n",
    "# we will develop our own class too."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = KNeighborsClassifier() # model is object of this inbuilt class."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "KNeighborsClassifier?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KNeighborsClassifier()"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# training the modelhappens in fit function\n",
    "# plotting the points on graph\n",
    "# In KNN algo., no work will happen in training knn model.\n",
    "model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 7, 0, 9, 4, 5, 4, 6, 9, 2], dtype=uint8)"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(X_test[0:10]) # predicting values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([7, 7, 0, 9, 4, 5, 4, 6, 9, 2], dtype=uint8)"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test[:10] # actual ones"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x25c242580a0>"
      ]
     },
     "execution_count": 89,
     "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": [
    "plt.imshow(X_test[0].reshape(28,28), cmap = 'gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.95"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.score(X_test[0:100], y_test[0:100])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Custom Implementation of KNN "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "class CustomKNN:\n",
    "    \n",
    "    # constructor\n",
    "    def __init__(self, n_neighbours = 5):\n",
    "        self.n_neighbours = n_neighbours\n",
    "    \n",
    "    # training function\n",
    "    def fit(self, X, y):\n",
    "        self._X = X.astype(np.int64)\n",
    "        self._y = y\n",
    "        \n",
    "   # predict the point\n",
    "   # Given a single point, tell me to which class it belong.\n",
    "    def predict_point(self, point):\n",
    "        \n",
    "        # storing the distance of given 'point' from each points in the training data set.\n",
    "        list_dist = []\n",
    "        \n",
    "        # these points are from our training data.\n",
    "        for x_point, y_point in zip(self._X, self._y):\n",
    "            dist_point = ((point - x_point) ** 2).sum()\n",
    "            list_dist.append([dist_point, y_point])\n",
    "        \n",
    "        # sorting the list according to distance.\n",
    "        sorted_dist = sorted(list_dist)\n",
    "            \n",
    "        # selecting top k neighbours.   \n",
    "        top_k = sorted_dist[:self.n_neighbours]\n",
    "    \n",
    "       # taking the counts.\n",
    "       # we are finfing unique elements according to classes to which point belong not the unique 'distances'.\n",
    "        items, counts = np.unique(np.array(top_k)[:, 1], return_counts = True)\n",
    "        ans = items[np.argmax (counts)]\n",
    "        return ans\n",
    "            \n",
    "   # predict function will give answer/output for each number in array.    \n",
    "    def predict(self, X):\n",
    "        results = []\n",
    "        for point in X:\n",
    "            results.append(self.predict_point(point))\n",
    "        return np.array(results, dtype = int)\n",
    "    \n",
    "   # Score to measure model's accuracy between its predicted and actual output.  \n",
    "    def score(self, X, y):\n",
    "        return sum(self.predict(X) == y)/ len(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "m2 = CustomKNN() # Initializing object of our custom class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "m2.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 7, 0, 9, 4, 5, 4, 6, 9, 2])"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m2.predict(X_test[:10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([7, 7, 0, 9, 4, 5, 4, 6, 9, 2], dtype=uint8)"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_test[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.95"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m2.score(X_test[:100], y_test[:100])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}