My_digit_recognition-checkpoint.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "7TbSMonR0TgW"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorflow\\python\\framework\\dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n",
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n"
     ]
    }
   ],
   "source": [
    "from tensorflow import keras as tfk"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "CbOMGO_Y0dx0"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 54
    },
    "colab_type": "code",
    "id": "vvVbN_gV0oSX",
    "outputId": "bb74b89e-b37c-43ec-fa0a-ca04608cad99"
   },
   "outputs": [],
   "source": [
    "mnist = tfk.datasets.mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "2Iwviejz0wt7",
    "outputId": "1a234fcd-3779-4fa8-d811-3e00c0532217"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(mnist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "C_eI28ng073n"
   },
   "outputs": [],
   "source": [
    "train, test = mnist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "8KcKHAqX1DSf",
    "outputId": "08f44df5-6d12-409f-c425-832eb2a4712b"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2, 2)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train), len(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ik_lk2gk1F_W"
   },
   "outputs": [],
   "source": [
    "X_train, y_train = train\n",
    "X_test, y_test = test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "qgYaJYnV1NmA",
    "outputId": "74159976-c9ab-4631-f4dd-720737359b01"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((60000, 28, 28), (60000,), (10000, 28, 28), (10000,))"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, y_train.shape, X_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "3ywuElsG1d4G",
    "outputId": "e510a793-7877-4d1f-9a5a-1c406280bc34"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "numpy.ndarray"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 283
    },
    "colab_type": "code",
    "id": "zCnTHyU49ZZN",
    "outputId": "4208ac34-faec-4cc2-d6a8-6841a941e4ea"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4\n"
     ]
    },
    {
     "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": [
    "# To see the image in our training data:\n",
    "\n",
    "plt.imshow(X_train[700], cmap=\"gray\")\n",
    "print(y_train[700])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "sVaiEtxy_AFK",
    "outputId": "68370b19-53b8-401c-c8c8-024e281f0efe"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0, 255)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train[800].min(), X_train[800].max() # We can see that the images are not normalized."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "sQlpv_8lCsik"
   },
   "outputs": [],
   "source": [
    "# Lets normalize the train and test image:\n",
    "\n",
    "X_train = X_train/255\n",
    "X_test = X_test/255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "zKT1sThlCBnJ",
    "outputId": "48295730-dffe-467c-d5a3-d58dc58b2545"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.0, 1.0)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train[800].min(), X_train[800].max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "UNLfiQdsCQ1l",
    "outputId": "4b8d789e-0136-492c-adb6-92533da216b7"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.0, 1.0)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test[100].min(), X_test[100].max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "aljaELrqDBcn"
   },
   "outputs": [],
   "source": [
    "# Both the train and test images are normalized now."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "r8FEpejuDGcl",
    "outputId": "3c85b5f3-2a18-4ce0-bb39-0c0093d2acc9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.38431373, 0.87843137, 0.16470588, 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.11764706, 0.44313725, 0.29019608,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.03137255, 0.81176471, 0.81960784, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.15294118, 0.82745098, 0.84313725, 0.11372549,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.42352941, 0.99607843, 0.4       , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.01176471, 0.66666667, 0.96862745, 0.18039216, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.04313725,\n",
       "        0.83529412, 0.89019608, 0.04705882, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.29411765, 0.99607843, 0.44313725, 0.22745098, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.54901961,\n",
       "        0.99607843, 0.31764706, 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.61960784, 0.9372549 , 0.36078431, 0.80392157, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.15294118, 0.98431373,\n",
       "        0.79607843, 0.02352941, 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.16078431,\n",
       "        0.98431373, 0.65098039, 0.85098039, 0.38431373, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.56862745, 0.99607843,\n",
       "        0.34509804, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.36862745,\n",
       "        0.99607843, 0.99607843, 0.94901961, 0.15686275, 0.        ,\n",
       "        0.        , 0.        , 0.17254902, 0.96470588, 0.71372549,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.52156863,\n",
       "        0.99607843, 0.99607843, 0.41960784, 0.        , 0.        ,\n",
       "        0.        , 0.01960784, 0.68627451, 0.94117647, 0.22352941,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.85098039,\n",
       "        0.99607843, 0.83137255, 0.07058824, 0.        , 0.        ,\n",
       "        0.        , 0.27843137, 0.99607843, 0.63137255, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.10980392, 0.98431373,\n",
       "        0.91372549, 0.18431373, 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.90980392, 0.8745098 , 0.03921569, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.2745098 , 0.99607843,\n",
       "        0.54117647, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.14901961, 0.98431373, 0.39215686, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.11764706, 0.98039216,\n",
       "        0.89803922, 0.41176471, 0.2745098 , 0.0745098 , 0.        ,\n",
       "        0.74509804, 0.82745098, 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.56470588,\n",
       "        0.99607843, 0.99607843, 0.99607843, 0.53333333, 0.3372549 ,\n",
       "        0.96862745, 0.45098039, 0.05098039, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.09019608,\n",
       "        0.4745098 , 0.54117647, 0.7372549 , 0.95686275, 0.99607843,\n",
       "        0.99607843, 0.98823529, 0.6       , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.54117647, 0.99607843,\n",
       "        0.68235294, 0.1372549 , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.01960784, 0.81960784, 0.85490196,\n",
       "        0.03529412, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.37254902, 0.99607843, 0.35686275,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.63529412, 0.99607843, 0.18039216,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.36470588, 0.87058824, 0.1254902 ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ]])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train[700]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 54
    },
    "colab_type": "code",
    "id": "x1SQbcbmDVll",
    "outputId": "caeccd6c-de62-46a7-b206-c76bd783de40"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "7\n",
      "2\n"
     ]
    }
   ],
   "source": [
    "print(y_train[5000])\n",
    "print(y_test[400])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "-eTT1y4nLQj7"
   },
   "outputs": [],
   "source": [
    "# Our y_train and y_test are still real no. values, we need to make them categorical.\n",
    "\n",
    "y_train = tfk.utils.to_categorical(y_train)\n",
    "y_test = tfk.utils.to_categorical(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 54
    },
    "colab_type": "code",
    "id": "1O3aHUVuLnG0",
    "outputId": "5f17aa82-aa58-4e7f-fa82-c8b754410769"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]\n",
      "[0. 0. 1. 0. 0. 0. 0. 0. 0. 0.]\n"
     ]
    }
   ],
   "source": [
    "print(y_train[5000])\n",
    "print(y_test[400])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "3rgpyKDQLpPu",
    "outputId": "e2eb19d4-61f8-42ec-ab60-c25a21718a3e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((60000, 10), (10000, 10))"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "woKtJ3ybL0JK"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\Admin\\anaconda3\\lib\\site-packages\\tensorflow\\python\\ops\\init_ops.py:1251: calling VarianceScaling.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Call initializer instance with the dtype argument instead of passing it to the constructor\n"
     ]
    }
   ],
   "source": [
    "# Lets start with model building now:\n",
    "\n",
    "model = tfk.models.Sequential()\n",
    "\n",
    "model.add(tfk.layers.InputLayer(input_shape=(28,28)))\n",
    "model.add(tfk.layers.Flatten())\n",
    "model.add(tfk.layers.Dense(128, activation='relu'))\n",
    "model.add(tfk.layers.Dense(128, activation = 'relu'))\n",
    "model.add(tfk.layers.Dense(10, activation='softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "8Bx8UbooPx4G"
   },
   "outputs": [],
   "source": [
    "model.compile(optimizer='rmsprop', loss=tfk.losses.categorical_crossentropy, metrics=['acc'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 308
    },
    "colab_type": "code",
    "id": "VZisXsKmQa_-",
    "outputId": "50d2fe0c-791e-4c49-f5bf-ecbd92898bb9"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "flatten (Flatten)            (None, 784)               0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 128)               100480    \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 128)               16512     \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 10)                1290      \n",
      "=================================================================\n",
      "Total params: 118,282\n",
      "Trainable params: 118,282\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "W9ZDp_MNPTvd",
    "outputId": "d0173761-d6d5-45e1-f0f1-8917fa358210"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 48000 samples, validate on 12000 samples\n",
      "Epoch 1/30\n",
      "48000/48000 [==============================] - 1s 30us/sample - loss: 0.6892 - acc: 0.8172 - val_loss: 0.3146 - val_acc: 0.9114\n",
      "Epoch 2/30\n",
      "48000/48000 [==============================] - 1s 27us/sample - loss: 0.2971 - acc: 0.9157 - val_loss: 0.2547 - val_acc: 0.9254\n",
      "Epoch 3/30\n",
      "48000/48000 [==============================] - 1s 27us/sample - loss: 0.2295 - acc: 0.9333 - val_loss: 0.2043 - val_acc: 0.9398\n",
      "Epoch 4/30\n",
      "48000/48000 [==============================] - 1s 27us/sample - loss: 0.1890 - acc: 0.9455 - val_loss: 0.1704 - val_acc: 0.9515\n",
      "Epoch 5/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.1564 - acc: 0.9549 - val_loss: 0.1566 - val_acc: 0.9544\n",
      "Epoch 6/30\n",
      "48000/48000 [==============================] - 1s 27us/sample - loss: 0.1373 - acc: 0.9596 - val_loss: 0.1485 - val_acc: 0.9574\n",
      "Epoch 7/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.1204 - acc: 0.9647 - val_loss: 0.1260 - val_acc: 0.9615\n",
      "Epoch 8/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.1031 - acc: 0.9691 - val_loss: 0.1262 - val_acc: 0.9607\n",
      "Epoch 9/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.0923 - acc: 0.9728 - val_loss: 0.1121 - val_acc: 0.9658\n",
      "Epoch 10/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.0821 - acc: 0.9762 - val_loss: 0.1056 - val_acc: 0.9683\n",
      "Epoch 11/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.0732 - acc: 0.9778 - val_loss: 0.1113 - val_acc: 0.9671\n",
      "Epoch 12/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.0658 - acc: 0.9801 - val_loss: 0.1035 - val_acc: 0.9684\n",
      "Epoch 13/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.0581 - acc: 0.9831 - val_loss: 0.0928 - val_acc: 0.9723\n",
      "Epoch 14/30\n",
      "48000/48000 [==============================] - 1s 29us/sample - loss: 0.0525 - acc: 0.9844 - val_loss: 0.1051 - val_acc: 0.9676\n",
      "Epoch 15/30\n",
      "48000/48000 [==============================] - 1s 29us/sample - loss: 0.0480 - acc: 0.9856 - val_loss: 0.0950 - val_acc: 0.9727\n",
      "Epoch 16/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.0425 - acc: 0.9877 - val_loss: 0.0896 - val_acc: 0.9747\n",
      "Epoch 17/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.0385 - acc: 0.9888 - val_loss: 0.1219 - val_acc: 0.9658\n",
      "Epoch 18/30\n",
      "48000/48000 [==============================] - 1s 30us/sample - loss: 0.0363 - acc: 0.9896 - val_loss: 0.0923 - val_acc: 0.9728\n",
      "Epoch 19/30\n",
      "48000/48000 [==============================] - 2s 36us/sample - loss: 0.0317 - acc: 0.9908 - val_loss: 0.0951 - val_acc: 0.9723\n",
      "Epoch 20/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.0283 - acc: 0.9917 - val_loss: 0.0851 - val_acc: 0.9768\n",
      "Epoch 21/30\n",
      "48000/48000 [==============================] - 1s 27us/sample - loss: 0.0260 - acc: 0.9928 - val_loss: 0.0866 - val_acc: 0.9757\n",
      "Epoch 22/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.0244 - acc: 0.9931 - val_loss: 0.0851 - val_acc: 0.9755\n",
      "Epoch 23/30\n",
      "48000/48000 [==============================] - 1s 29us/sample - loss: 0.0210 - acc: 0.9943 - val_loss: 0.1169 - val_acc: 0.9668\n",
      "Epoch 24/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.0206 - acc: 0.9940 - val_loss: 0.0925 - val_acc: 0.9733\n",
      "Epoch 25/30\n",
      "48000/48000 [==============================] - 2s 34us/sample - loss: 0.0184 - acc: 0.9952 - val_loss: 0.0927 - val_acc: 0.9752\n",
      "Epoch 26/30\n",
      "48000/48000 [==============================] - 1s 27us/sample - loss: 0.0157 - acc: 0.9958 - val_loss: 0.0981 - val_acc: 0.9738\n",
      "Epoch 27/30\n",
      "48000/48000 [==============================] - 1s 27us/sample - loss: 0.0149 - acc: 0.9964 - val_loss: 0.0877 - val_acc: 0.9762\n",
      "Epoch 28/30\n",
      "48000/48000 [==============================] - 1s 28us/sample - loss: 0.0134 - acc: 0.9962 - val_loss: 0.0983 - val_acc: 0.9744\n",
      "Epoch 29/30\n",
      "48000/48000 [==============================] - 1s 27us/sample - loss: 0.0099 - acc: 0.9977 - val_loss: 0.1003 - val_acc: 0.9753\n",
      "Epoch 30/30\n",
      "48000/48000 [==============================] - 2s 33us/sample - loss: 0.0111 - acc: 0.9973 - val_loss: 0.0948 - val_acc: 0.9751\n"
     ]
    }
   ],
   "source": [
    "model_fit = model.fit(x=X_train, y= y_train, batch_size=1000,epochs=30, validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "I5azmvX4FKQh"
   },
   "outputs": [],
   "source": [
    "model.save('C:\\\\Users\\\\Admin\\\\Desktop\\\\Handwritten digit MNIST\\\\Handwriting.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "gFknCFRcRVBr",
    "outputId": "5def4cf4-d4b5-48bc-b468-99cc6a9eb415"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['loss', 'acc', 'val_loss', 'val_acc'])"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_fit.history.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 563
    },
    "colab_type": "code",
    "id": "BUiEgl4tSEF7",
    "outputId": "ae8a95df-d3fd-4d04-c8f3-b7a96506122b"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.6891920777658621,\n",
       " 0.2971179373562336,\n",
       " 0.2295037896061937,\n",
       " 0.18902780146648487,\n",
       " 0.15644852413485447,\n",
       " 0.13726038935904702,\n",
       " 0.12041077095394333,\n",
       " 0.10310267610475421,\n",
       " 0.09228029109848042,\n",
       " 0.08214078478825589,\n",
       " 0.0732337834003071,\n",
       " 0.06579765692974131,\n",
       " 0.058081653357173,\n",
       " 0.052529982601602875,\n",
       " 0.0480190699842448,\n",
       " 0.04254512535408139,\n",
       " 0.03852618505091717,\n",
       " 0.03630813560448587,\n",
       " 0.03168434308220943,\n",
       " 0.02830335278607284,\n",
       " 0.02598919744680946,\n",
       " 0.024446504112953942,\n",
       " 0.0209994925535284,\n",
       " 0.020582258216260623,\n",
       " 0.01842309096052001,\n",
       " 0.015695518425976236,\n",
       " 0.014942035981221125,\n",
       " 0.01337523219990544,\n",
       " 0.0099326836449715,\n",
       " 0.011095349706010893]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_fit.history['loss']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 283
    },
    "colab_type": "code",
    "id": "ZzbqQOnVSKSz",
    "outputId": "a7891131-5929-41e1-d61a-8928c3175b67"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1f29b5c6688>"
      ]
     },
     "execution_count": 28,
     "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": [
    "# Plotting graph of # of epochs Vs. Training loss/validation loss:\n",
    "\n",
    "plt.plot(model_fit.epoch,model_fit.history['loss'],label=\"training_loss\")\n",
    "plt.plot(model_fit.history['val_loss'], label=\"validation_loss\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 283
    },
    "colab_type": "code",
    "id": "Ntt0XT0lSkWV",
    "outputId": "11d77a82-b52e-4e72-9bff-c3ae90095da2"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1f29b653f88>"
      ]
     },
     "execution_count": 29,
     "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": [
    "# Plotting graph of # of epochs Vs. Training acc/validation acc:\n",
    "\n",
    "plt.plot(model_fit.epoch,model_fit.history['acc'],label=\"training_acc\")\n",
    "plt.plot(model_fit.history['val_acc'], label=\"validation_acc\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 54
    },
    "colab_type": "code",
    "id": "m1u7tsnoVS28",
    "outputId": "46bc64db-cd81-48a8-f056-24f15e40bc7d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000/10000 [==============================] - 0s 42us/sample - loss: 0.0810 - acc: 0.9787\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.08098559966890607, 0.9787]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 72
    },
    "colab_type": "code",
    "id": "lgAcksOuWPS8",
    "outputId": "a533b5d8-9eb0-44d5-bf37-50b1b896e979"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.7702431e-04, 3.8700912e-07, 1.1449781e-05, 1.2131173e-09,\n",
       "        9.9975461e-01, 8.6645359e-06, 1.9441104e-05, 8.6734972e-06,\n",
       "        5.4045645e-06, 1.4278308e-05]], dtype=float32)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(X_train[700].reshape(1,28,28))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "QU9HuBJzWlnK"
   },
   "outputs": [],
   "source": [
    "# The above shows the probability for each digit it might be.!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "6LfI-k4oXKUf",
    "outputId": "22571b6c-66fe-4536-a93a-d9c7f67d8ec2"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argmax(model.predict(X_train[700].reshape(1,28,28)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "8U0lfJlKX6sm",
    "outputId": "e1768d6b-42d0-4a3e-bea1-6d0655c84a2b"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([7, 2, 1, ..., 4, 5, 6], dtype=int64)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argmax(model.predict(X_test), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 72
    },
    "colab_type": "code",
    "id": "sHeT_8tiXThp",
    "outputId": "32885553-e730-4b9c-dc80-6aa1aeeeade2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "JPEG\n",
      "(350, 406)\n",
      "RGB\n"
     ]
    }
   ],
   "source": [
    "# Lets us try and upload our handwritten digit to classify it: \n",
    "\n",
    "from PIL import Image\n",
    "import numpy as np\n",
    "\n",
    "img = Image.open('8_paint.jpg')\n",
    "print(img.format)\n",
    "print(img.size)\n",
    "print(img.mode)\n",
    "\n",
    "img.show()\n",
    "\n",
    "# Note: we need some additional package for show() to work as PIL doesn't seem to use the default image viewer. It should but does not."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 305
    },
    "colab_type": "code",
    "id": "3tBSqgZEUTX4",
    "outputId": "90c25dc2-9246-4e15-dcb5-abc25ebcf477"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "uint8\n",
      "(406, 350, 3)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Lets us try the same and upload our handwritten digit using MATPLOTLIB:\n",
    "\n",
    "from matplotlib import image\n",
    "from matplotlib import pyplot\n",
    "\n",
    "img = image.imread('8_handwritten.jpg')\n",
    "print(img.dtype)\n",
    "print(img.shape)\n",
    "pyplot.imshow(img)\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]],\n",
       "\n",
       "       [[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]],\n",
       "\n",
       "       [[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]],\n",
       "\n",
       "       ...,\n",
       "\n",
       "       [[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]],\n",
       "\n",
       "       [[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]],\n",
       "\n",
       "       [[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]]], dtype=uint8)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "tzaEzxDwYdGW"
   },
   "source": [
    "##### However, In Python, Pillow is the most popular and standard library when it comes to working with image data.\n",
    "###### So, we will be working with Pillow library for further manipulation and processing.!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 54
    },
    "colab_type": "code",
    "id": "UevH7Z7RaCZO",
    "outputId": "dffa68c9-30f7-445d-c36a-33c14ac15ee0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'numpy.ndarray'>\n",
      "(406, 350, 3)\n"
     ]
    }
   ],
   "source": [
    "# Converting image to a numpy array:\n",
    "\n",
    "from numpy import asarray\n",
    "\n",
    "img_array = asarray(img)\n",
    "print(type(img_array))\n",
    "print(img_array.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 908
    },
    "colab_type": "code",
    "id": "TIGW3d4KaVPw",
    "outputId": "d73c7fd8-c259-4844-b2a5-a9a4becbd03a"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]],\n",
       "\n",
       "       [[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]],\n",
       "\n",
       "       [[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]],\n",
       "\n",
       "       ...,\n",
       "\n",
       "       [[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]],\n",
       "\n",
       "       [[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]],\n",
       "\n",
       "       [[255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        ...,\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255],\n",
       "        [255, 255, 255]]], dtype=uint8)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img_array  # pixels of image stored in ndarray"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 72
    },
    "colab_type": "code",
    "id": "3MqKQMxpaq9j",
    "outputId": "df8ecfbf-8983-4e8c-b2df-3b094f3f2a21"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'PIL.Image.Image'>\n",
      "RGB\n",
      "(350, 406)\n"
     ]
    }
   ],
   "source": [
    "# To create the Pillow image from ndarray:\n",
    "\n",
    "pill_img = Image.fromarray(img_array)\n",
    "print(type(pill_img))\n",
    "print(pill_img.mode) \n",
    "print(pill_img.size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 54
    },
    "colab_type": "code",
    "id": "y4rw-XmbbXKf",
    "outputId": "587a985e-6708-48eb-8e7f-073d1e1f1518"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'numpy.ndarray'>\n",
      "(406, 350)\n"
     ]
    }
   ],
   "source": [
    "# Manipulating the image:\n",
    "\n",
    "im = np.array(Image.open('8_handwritten.jpg').convert('L')) \n",
    "print(type(im))\n",
    "print(im.shape)\n",
    "gray_im= Image.fromarray(im).save('8_pil.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 145
    },
    "colab_type": "code",
    "id": "6L7GNZXVGjfs",
    "outputId": "0782c372-99b4-4b82-faa2-0c4a10d0cdaa"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[255, 255, 255, ..., 255, 255, 255],\n",
       "       [255, 255, 255, ..., 255, 255, 255],\n",
       "       [255, 255, 255, ..., 255, 255, 255],\n",
       "       ...,\n",
       "       [255, 255, 255, ..., 255, 255, 255],\n",
       "       [255, 255, 255, ..., 255, 255, 255],\n",
       "       [255, 255, 255, ..., 255, 255, 255]], dtype=uint8)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "im"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Gpom0iKUGcex"
   },
   "outputs": [],
   "source": [
    "im = np.invert(im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 145
    },
    "colab_type": "code",
    "id": "GXzKYgFiI2Fv",
    "outputId": "4e1baa09-13a9-4099-b832-b611c9705875"
   },
   "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": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "im"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "0Ot-_64KGs45"
   },
   "outputs": [],
   "source": [
    "invert_im = Image.fromarray(im).save('8_inverted.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "5b0orLwUcHZe"
   },
   "outputs": [],
   "source": [
    "# Open the TEST image, resizing the Image and converting to a grayscale:\n",
    "\n",
    "from PIL import Image\n",
    "from numpy import asarray\n",
    "\n",
    "im = np.array(Image.open('8_inverted.png').resize((28,28)))\n",
    "Image.fromarray(im).save('resized_8.png')\n",
    "#print(\"After resizing:\",img_resized.shape)\n",
    "#im = np.array(Image.open('resized_1.jpg').convert('L')) #you can pass multiple arguments in single line\n",
    "#print(type(im))\n",
    "#gray_im= Image.fromarray(im).save('1_final.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "Dj9kRqpHlBRE",
    "outputId": "6545cc6a-8542-4fcb-f57a-d928d737fdf0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   5,  34,  54,\n",
       "         97,  91,  92,  57,   2,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,  64, 153, 158, 125,\n",
       "        103,  99, 118, 184, 158,   9,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,   6,  73,  90,  27,   3,   0,\n",
       "          0,   0,   0,   5, 119, 137,   3,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   1,  85,  59,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,  97,  91,   1,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,  70,  45,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,  18, 155,   9,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,  15,  89,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,  32, 140,   4,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,  45,  76,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,  97,  75,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,  21, 124,   7,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,  44, 151,  11,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,  71, 141,  28,   0,   0,   0,   0,\n",
       "          0,   0,   0,  16, 164,  64,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,  70, 166, 105,  48,   3,   0,\n",
       "          0,   0,  35, 150,  58,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,   0,   7,  67, 122, 134, 143,\n",
       "         69,  31, 143,  66,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   4,  67,\n",
       "        217, 211, 119,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   8,  90, 123,\n",
       "         85,  33, 126, 123,  33,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,  20, 127,  99,  12,\n",
       "          0,   0,   0,  56, 159,  53,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,   0,  59, 103,  43,   0,   0,\n",
       "          0,   0,   0,   0,  28, 160,  27,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   2,  96,  97,   7,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,  61, 127,   5,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,  67,  92,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,   0, 163,  38,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   2, 102,   7,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,   0, 142,  50,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   3,  85,   2,   0,   1,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,   0, 132,  50,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   7,  87,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,   0, 118,  53,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,  13, 116,   1,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,   0, 111,  37,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   1,  98,  57,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   8,  67,  60,   3,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   1, 101,  49,   0,   0,   0,   0,\n",
       "          0,   0,   0,  63, 157,  46,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,  16, 105, 113,  70,  36,  33,\n",
       "         32,  37, 111, 150,  46,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,  50,  90,  98, 123,\n",
       "        122,  95,  53,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0],\n",
       "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
       "          0,   0]], dtype=uint8)"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "im"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "RXgOXHiqAHPd"
   },
   "outputs": [],
   "source": [
    "im = im/255.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "lanLSJ1QfGth",
    "outputId": "eecd8532-c8b3-4eb3-8955-e24db216c94e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(28, 28)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "im.shape # now the im has (28,28,1) shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "_6EfajtCHGzD",
    "outputId": "4bfc1308-fa27-406a-ea0b-ad977b157f2a"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.01960784, 0.13333333, 0.21176471, 0.38039216, 0.35686275,\n",
       "        0.36078431, 0.22352941, 0.00784314, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.25098039,\n",
       "        0.6       , 0.61960784, 0.49019608, 0.40392157, 0.38823529,\n",
       "        0.4627451 , 0.72156863, 0.61960784, 0.03529412, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.02352941, 0.28627451, 0.35294118,\n",
       "        0.10588235, 0.01176471, 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.01960784, 0.46666667, 0.5372549 , 0.01176471,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.00392157, 0.33333333, 0.23137255, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.38039216, 0.35686275,\n",
       "        0.00392157, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.2745098 , 0.17647059, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.07058824, 0.60784314,\n",
       "        0.03529412, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.05882353, 0.34901961, 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.1254902 , 0.54901961,\n",
       "        0.01568627, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.17647059, 0.29803922, 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.38039216, 0.29411765,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.08235294, 0.48627451, 0.02745098, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.17254902, 0.59215686, 0.04313725,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.27843137, 0.55294118, 0.10980392, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.0627451 , 0.64313725, 0.25098039, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.2745098 , 0.65098039, 0.41176471,\n",
       "        0.18823529, 0.01176471, 0.        , 0.        , 0.        ,\n",
       "        0.1372549 , 0.58823529, 0.22745098, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.02745098, 0.2627451 ,\n",
       "        0.47843137, 0.5254902 , 0.56078431, 0.27058824, 0.12156863,\n",
       "        0.56078431, 0.25882353, 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.01568627, 0.2627451 , 0.85098039, 0.82745098,\n",
       "        0.46666667, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.03137255, 0.35294118, 0.48235294, 0.33333333, 0.12941176,\n",
       "        0.49411765, 0.48235294, 0.12941176, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.07843137,\n",
       "        0.49803922, 0.38823529, 0.04705882, 0.        , 0.        ,\n",
       "        0.        , 0.21960784, 0.62352941, 0.20784314, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.23137255, 0.40392157,\n",
       "        0.16862745, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.10980392, 0.62745098, 0.10588235,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.00784314, 0.37647059, 0.38039216, 0.02745098,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.23921569, 0.49803922,\n",
       "        0.01960784, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.2627451 , 0.36078431, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.63921569,\n",
       "        0.14901961, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.00784314, 0.4       , 0.02745098, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.55686275,\n",
       "        0.19607843, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.01176471, 0.33333333, 0.00784314, 0.        , 0.00392157,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.51764706,\n",
       "        0.19607843, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.02745098, 0.34117647, 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.4627451 ,\n",
       "        0.20784314, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.05098039, 0.45490196, 0.00392157, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.43529412,\n",
       "        0.14509804, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.00392157, 0.38431373, 0.22352941, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.03137255, 0.2627451 , 0.23529412,\n",
       "        0.01176471, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.00392157, 0.39607843, 0.19215686, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.24705882, 0.61568627, 0.18039216, 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.0627451 , 0.41176471, 0.44313725,\n",
       "        0.2745098 , 0.14117647, 0.12941176, 0.1254902 , 0.14509804,\n",
       "        0.43529412, 0.58823529, 0.18039216, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.19607843,\n",
       "        0.35294118, 0.38431373, 0.48235294, 0.47843137, 0.37254902,\n",
       "        0.20784314, 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ]])"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "im"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 72
    },
    "colab_type": "code",
    "id": "K4Tla9-RHhDa",
    "outputId": "ec37d7d2-a47b-44c7-a731-48cd792add39"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[8.2334582e-06, 4.6577598e-03, 1.2507220e-03, 1.1941760e-01,\n",
       "        4.0326908e-05, 4.6668015e-03, 4.7576606e-02, 4.5850541e-04,\n",
       "        8.2190287e-01, 2.0613496e-05]], dtype=float32)"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(im.reshape(1,28,28))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "6jXoFZphfMQq",
    "outputId": "2ac620ba-665f-4bb5-ccd6-9bac296ff02a"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argmax(model.predict(im.reshape(1,28,28)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "WczwUKroJDBG"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "accelerator": "TPU",
  "colab": {
   "collapsed_sections": [],
   "name": "My_digit_recognition (1).ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}