2_MNIST_Fashion_Classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "jYysdyb-CaWM"
   },
   "source": [
    "# Train your first neural network: basic classification\n",
    "\n",
    "Adapted from TensorFlow example, Copyright 2018 The TensorFlow Authors, 2017 François Chollet"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "FbVhjPpzn6BM"
   },
   "source": [
    "This guide trains a neural network model to classify images of clothing, like sneakers and shirts. It's okay if you don't understand all the details, this is a fast-paced overview of a complete TensorFlow program with the details explained as we go.\n",
    "\n",
    "This guide uses [tf.keras](https://www.tensorflow.org/guide/keras), a high-level API to build and train models in TensorFlow."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "dzLKpmZICaWN"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.12.0\n"
     ]
    }
   ],
   "source": [
    "from __future__ import absolute_import, division, print_function\n",
    "\n",
    "# TensorFlow and tf.keras\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "\n",
    "# Helper libraries\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "print(tf.__version__)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "yR0EdgrLCaWR"
   },
   "source": [
    "## Import the Fashion MNIST dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "DLdCchMdCaWQ"
   },
   "source": [
    "This guide uses the [Fashion MNIST](https://github.com/zalandoresearch/fashion-mnist) dataset which contains 70,000 grayscale images in 10 categories. The images show individual articles of clothing at low resolution (28 by 28 pixels), as seen here:\n",
    "\n",
    "<table>\n",
    "  <tr><td>\n",
    "    <img src=\"https://tensorflow.org/images/fashion-mnist-sprite.png\"\n",
    "         alt=\"Fashion MNIST sprite\"  width=\"600\">\n",
    "  </td></tr>\n",
    "  <tr><td align=\"center\">\n",
    "    <b>Figure 1.</b> <a href=\"https://github.com/zalandoresearch/fashion-mnist\">Fashion-MNIST samples</a> (by Zalando, MIT License).<br/>&nbsp;\n",
    "  </td></tr>\n",
    "</table>\n",
    "\n",
    "Fashion MNIST is intended as a drop-in replacement for the classic [MNIST](http://yann.lecun.com/exdb/mnist/) dataset—often used as the \"Hello, World\" of machine learning programs for computer vision. The MNIST dataset contains images of handwritten digits (0, 1, 2, etc) in an identical format to the articles of clothing we'll use here.\n",
    "\n",
    "This guide uses Fashion MNIST for variety, and because it's a slightly more challenging problem than regular MNIST. Both datasets are relatively small and are used to verify that an algorithm works as expected. They're good starting points to test and debug code. \n",
    "\n",
    "We will use 60,000 images to train the network and 10,000 images to evaluate how accurately the network learned to classify images. You can access the Fashion MNIST directly from TensorFlow, just import and load the data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "7MqDQO0KCaWS"
   },
   "outputs": [],
   "source": [
    "fashion_mnist = keras.datasets.fashion_mnist\n",
    "# fashion_mnist = keras.datasets.mnist\n",
    "\n",
    "(train_images, train_labels), (test_images, test_labels) = fashion_mnist.load_data()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "t9FDsUlxCaWW"
   },
   "source": [
    "Loading the dataset returns four NumPy arrays:\n",
    "\n",
    "* The `train_images` and `train_labels` arrays are the *training set*—the data the model uses to learn.\n",
    "* The model is tested against the *test set*, the `test_images`, and `test_labels` arrays.\n",
    "\n",
    "The images are 28x28 NumPy arrays, with pixel values ranging between 0 and 255. The *labels* are an array of integers, ranging from 0 to 9. These correspond to the *class* of clothing the image represents:\n",
    "\n",
    "<table>\n",
    "  <tr>\n",
    "    <th>Label</th>\n",
    "    <th>Class</th> \n",
    "  </tr>\n",
    "  <tr>\n",
    "    <td>0</td>\n",
    "    <td>T-shirt/top</td> \n",
    "  </tr>\n",
    "  <tr>\n",
    "    <td>1</td>\n",
    "    <td>Trouser</td> \n",
    "  </tr>\n",
    "    <tr>\n",
    "    <td>2</td>\n",
    "    <td>Pullover</td> \n",
    "  </tr>\n",
    "    <tr>\n",
    "    <td>3</td>\n",
    "    <td>Dress</td> \n",
    "  </tr>\n",
    "    <tr>\n",
    "    <td>4</td>\n",
    "    <td>Coat</td> \n",
    "  </tr>\n",
    "    <tr>\n",
    "    <td>5</td>\n",
    "    <td>Sandal</td> \n",
    "  </tr>\n",
    "    <tr>\n",
    "    <td>6</td>\n",
    "    <td>Shirt</td> \n",
    "  </tr>\n",
    "    <tr>\n",
    "    <td>7</td>\n",
    "    <td>Sneaker</td> \n",
    "  </tr>\n",
    "    <tr>\n",
    "    <td>8</td>\n",
    "    <td>Bag</td> \n",
    "  </tr>\n",
    "    <tr>\n",
    "    <td>9</td>\n",
    "    <td>Ankle boot</td> \n",
    "  </tr>\n",
    "</table>\n",
    "\n",
    "Each image is mapped to a single label. Since the *class names* are not included with the dataset, store them here to use later when plotting the images:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "IjnLH5S2CaWx"
   },
   "outputs": [],
   "source": [
    "class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat', \n",
    "               'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Brm0b_KACaWX"
   },
   "source": [
    "## Explore the data\n",
    "\n",
    "Let's explore the format of the dataset before training the model. The following shows there are 60,000 images in the training set, with each image represented as 28 x 28 pixels:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "zW5k_xz1CaWX"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 28, 28)"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_images.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "cIAcvQqMCaWf"
   },
   "source": [
    "Likewise, there are 60,000 labels in the training set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "TRFYHB2mCaWb"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "60000"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "YSlYxFuRCaWk"
   },
   "source": [
    "Each label is an integer between 0 and 9:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "XKnCTHz4CaWg"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([9, 0, 0, ..., 3, 0, 5], dtype=uint8)"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "TMPI88iZpO2T"
   },
   "source": [
    "There are 10,000 images in the test set. Again, each image is represented as 28 x 28 pixels:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "2KFnYlcwCaWl"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 28, 28)"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_images.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "rd0A0Iu0CaWq"
   },
   "source": [
    "And the test set contains 10,000 images labels:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "iJmPr5-ACaWn"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10000"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(test_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ES6uQoLKCaWr"
   },
   "source": [
    "## Preprocess the data\n",
    "\n",
    "The data must be preprocessed before training the network. If you inspect the first image in the training set, you will see that the pixel values fall in the range of 0 to 255:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "m4VEw8Ud9Quh"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "plt.imshow(train_images[0])\n",
    "plt.colorbar()\n",
    "plt.grid(False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Wz7l27Lz9S1P"
   },
   "source": [
    "We scale these values to a range of 0 to 1 before feeding to the neural network model. For this, we divide the values by 255. It's important that the *training set* and the *testing set* are preprocessed in the same way:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "bW5WzIPlCaWv"
   },
   "outputs": [],
   "source": [
    "train_images = train_images / 255.0\n",
    "\n",
    "test_images = test_images / 255.0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Ee638AlnCaWz"
   },
   "source": [
    "Display the first 25 images from the *training set* and display the class name below each image. Verify that the data is in the correct format and we're ready to build and train the network."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "oZTImqg_CaW1"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 25 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,10))\n",
    "for i in range(25):\n",
    "    plt.subplot(5,5,i+1)\n",
    "    plt.xticks([])\n",
    "    plt.yticks([])\n",
    "    plt.grid(False)\n",
    "    plt.imshow(train_images[i], cmap=plt.cm.binary)\n",
    "    plt.xlabel(class_names[train_labels[i]])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "59veuiEZCaW4"
   },
   "source": [
    "## Build the model\n",
    "\n",
    "Building the neural network requires configuring the layers of the model, then compiling the model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Gxg1XGm0eOBy"
   },
   "source": [
    "### Setup the layers\n",
    "\n",
    "The basic building block of a neural network is the *layer*. Layers extract representations from the data fed into them. And, hopefully, these representations are more meaningful for the problem at hand.\n",
    "\n",
    "Most of deep learning consists of chaining together simple layers. Most layers, like `tf.keras.layers.Dense`, have parameters that are learned during training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "9ODch-OFCaW4"
   },
   "outputs": [],
   "source": [
    "model = keras.Sequential([\n",
    "    keras.layers.Flatten(input_shape=(28, 28)),\n",
    "    keras.layers.Dense(128, activation=tf.nn.relu),\n",
    "    keras.layers.Dense(10, activation=tf.nn.softmax)\n",
    "])\n",
    "\n",
    "# experimenting with a smaller dense hidden layer\n",
    "# model = keras.Sequential([\n",
    "#     keras.layers.Flatten(input_shape=(28, 28)),\n",
    "#     keras.layers.Dense(32, activation=tf.nn.relu),\n",
    "#     keras.layers.Dense(10, activation=tf.nn.softmax)\n",
    "# ])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "gut8A_7rCaW6"
   },
   "source": [
    "The first layer in this network, `tf.keras.layers.Flatten`, transforms the format of the images from a 2d-array (of 28 by 28 pixels), to a 1d-array of 28 * 28 = 784 pixels. Think of this layer as unstacking rows of pixels in the image and lining them up. This layer has no parameters to learn; it only reformats the data.\n",
    "\n",
    "After the pixels are flattened, the network consists of a sequence of two `tf.keras.layers.Dense` layers. These are densely-connected, or fully-connected, neural layers. The first `Dense` layer has 128 nodes (or neurons). The second (and last) layer is a 10-node *softmax* layer—this returns an array of 10 probability scores that sum to 1. Each node contains a score that indicates the probability that the current image belongs to one of the 10 classes.\n",
    "\n",
    "### Compile the model\n",
    "\n",
    "Before the model is ready for training, it needs a few more settings. These are added during the model's *compile* step:\n",
    "\n",
    "* *Loss function* —This measures how accurate the model is during training. We want to minimize this function to \"steer\" the model in the right direction.\n",
    "* *Optimizer* —This is how the model is updated based on the data it sees and its loss function.\n",
    "* *Metrics* —Used to monitor the training and testing steps. The following example uses *accuracy*, the fraction of the images that are correctly classified."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Lhan11blCaW7"
   },
   "outputs": [],
   "source": [
    "model.compile(optimizer='adam', \n",
    "              loss='sparse_categorical_crossentropy',\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "qKF6uW-BCaW-"
   },
   "source": [
    "## Train the model\n",
    "\n",
    "Training the neural network model requires the following steps:\n",
    "\n",
    "1. Feed the training data to the model—in this example, the `train_images` and `train_labels` arrays.\n",
    "2. The model learns to associate images and labels.\n",
    "3. We ask the model to make predictions about a test set—in this example, the `test_images` array. We verify that the predictions match the labels from the `test_labels` array. \n",
    "\n",
    "To start training,  call the `model.fit` method—the model is \"fit\" to the training data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "xvwvpA64CaW_"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "60000/60000 [==============================] - 8s 129us/step - loss: 0.4976 - acc: 0.8249\n",
      "Epoch 2/5\n",
      "60000/60000 [==============================] - 7s 124us/step - loss: 0.3722 - acc: 0.8652\n",
      "Epoch 3/5\n",
      "60000/60000 [==============================] - 8s 126us/step - loss: 0.3360 - acc: 0.8757\n",
      "Epoch 4/5\n",
      "60000/60000 [==============================] - 8s 139us/step - loss: 0.3140 - acc: 0.8830\n",
      "Epoch 5/5\n",
      "60000/60000 [==============================] - 7s 117us/step - loss: 0.2953 - acc: 0.8906\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x7f5b67c32128>"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# tensorboard feedback\n",
    "session_name = \"./logs\"\n",
    "\n",
    "tbCallBack = keras.callbacks.TensorBoard(log_dir=session_name, histogram_freq=0, write_graph=True)\n",
    "model.fit(train_images, train_labels, epochs=5, callbacks=[tbCallBack])\n",
    "\n",
    "# tensorboard with histograms\n",
    "#tbCallBack = keras.callbacks.TensorBoard(log_dir=session_name, histogram_freq=1, write_graph=True, write_grads=True, write_images=True)\n",
    "# model.fit(train_images, train_labels, epochs=5, callbacks=[tbCallBack], validation_split=0.1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "W3ZVOhugCaXA"
   },
   "source": [
    "As the model trains, the loss and accuracy metrics are displayed. This model reaches an accuracy of about 0.88 (or 88%) on the training data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "oEw4bZgGCaXB"
   },
   "source": [
    "## Evaluate accuracy\n",
    "\n",
    "Next, compare how the model performs on the test dataset:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "VflXLEeECaXC"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000/10000 [==============================] - 1s 98us/step\n",
      "Test accuracy: 0.8778\n"
     ]
    }
   ],
   "source": [
    "test_loss, test_acc = model.evaluate(test_images, test_labels)\n",
    "\n",
    "print('Test accuracy:', test_acc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "yWfgsmVXCaXG"
   },
   "source": [
    "It turns out, the accuracy on the test dataset is a little less than the accuracy on the training dataset. This gap between training accuracy and test accuracy is an example of *overfitting*. Overfitting is when a machine learning model performs worse on new data than on their training data. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "xsoS7CPDCaXH"
   },
   "source": [
    "## Make predictions\n",
    "\n",
    "With the model trained, we can use it to make predictions about some images."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Gl91RPhdCaXI"
   },
   "outputs": [],
   "source": [
    "predictions = model.predict(test_images)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "x9Kk1voUCaXJ"
   },
   "source": [
    "Here, the model has predicted the label for each image in the testing set. Let's take a look at the first prediction:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "3DmJEUinCaXK"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2.4201192e-08, 5.5426010e-09, 1.5154046e-05, 1.7383365e-05,\n",
       "       2.0669425e-10, 2.3164923e-09, 4.1760408e-13, 9.9996698e-01,\n",
       "       7.5990108e-08, 3.8982421e-07], dtype=float32)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "-hw1hgeSCaXN"
   },
   "source": [
    "A prediction is an array of 10 numbers. These describe the \"confidence\" of the model that the image corresponds to each of the 10 different articles of clothing. We can see which label has the highest confidence value:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "qsqenuPnCaXO"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argmax(predictions[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "E51yS7iCCaXO"
   },
   "source": [
    "So the model is most confident that this image is an ankle boot, or `class_names[9]`. And we can check the test label to see this is correct:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Sd7Pgsu6CaXP"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_labels[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ygh2yYC972ne"
   },
   "source": [
    "We can graph this to look at the full set of 10 channels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "DvYmmrpIy6Y1"
   },
   "outputs": [],
   "source": [
    "def plot_image(i, predictions_array, true_label, img):\n",
    "  predictions_array, true_label, img = predictions_array[i], true_label[i], img[i]\n",
    "  plt.grid(False)\n",
    "  plt.xticks([])\n",
    "  plt.yticks([])\n",
    "  \n",
    "  plt.imshow(img, cmap=plt.cm.binary)\n",
    "\n",
    "  predicted_label = np.argmax(predictions_array)\n",
    "  if predicted_label == true_label:\n",
    "    color = 'blue'\n",
    "  else:\n",
    "    color = 'red'\n",
    "  \n",
    "  plt.xlabel(\"{} {:2.0f}% ({})\".format(class_names[predicted_label],\n",
    "                                100*np.max(predictions_array),\n",
    "                                class_names[true_label]),\n",
    "                                color=color)\n",
    "\n",
    "def plot_value_array(i, predictions_array, true_label):\n",
    "  predictions_array, true_label = predictions_array[i], true_label[i]\n",
    "  plt.grid(False)\n",
    "  plt.xticks([])\n",
    "  plt.yticks([])\n",
    "  thisplot = plt.bar(range(10), predictions_array, color=\"#777777\")\n",
    "  plt.ylim([0, 1]) \n",
    "  predicted_label = np.argmax(predictions_array)\n",
    " \n",
    "  thisplot[predicted_label].set_color('red')\n",
    "  thisplot[true_label].set_color('blue')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "d4Ov9OFDMmOD"
   },
   "source": [
    "Let's look at the 0th image, predictions, and prediction array. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "HV5jw-5HwSmO"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "i = 0\n",
    "plt.figure(figsize=(6,3))\n",
    "plt.subplot(1,2,1)\n",
    "plot_image(i, predictions, test_labels, test_images)\n",
    "plt.subplot(1,2,2)\n",
    "plot_value_array(i, predictions,  test_labels)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Ko-uzOufSCSe"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "i = 12\n",
    "plt.figure(figsize=(6,3))\n",
    "plt.subplot(1,2,1)\n",
    "plot_image(i, predictions, test_labels, test_images)\n",
    "plt.subplot(1,2,2)\n",
    "plot_value_array(i, predictions,  test_labels)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "kgdvGD52CaXR"
   },
   "source": [
    "Let's plot several images with their predictions. Correct prediction labels are blue and incorrect prediction labels are red. The number gives the percent (out of 100) for the predicted label. Note that it can be wrong even when very confident. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "hQlnbqaw2Qu_"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 30 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the first X test images, their predicted label, and the true label\n",
    "# Color correct predictions in blue, incorrect predictions in red\n",
    "num_rows = 5\n",
    "num_cols = 3\n",
    "num_images = num_rows*num_cols\n",
    "plt.figure(figsize=(2*2*num_cols, 2*num_rows))\n",
    "for i in range(num_images):\n",
    "  plt.subplot(num_rows, 2*num_cols, 2*i+1)\n",
    "  plot_image(i, predictions, test_labels, test_images)\n",
    "  plt.subplot(num_rows, 2*num_cols, 2*i+2)\n",
    "  plot_value_array(i, predictions, test_labels)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "R32zteKHCaXT"
   },
   "source": [
    "Finally, use the trained model to make a prediction about a single image. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "yRJ7JU7JCaXT"
   },
   "outputs": [],
   "source": [
    "# Grab an image from the test dataset\n",
    "img = test_images[0]\n",
    "\n",
    "print(img.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "vz3bVp21CaXV"
   },
   "source": [
    "`tf.keras` models are optimized to make predictions on a *batch*, or collection, of examples at once. So even though we're using a single image, we need to add it to a list:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "lDFh5yF_CaXW"
   },
   "outputs": [],
   "source": [
    "# Add the image to a batch where it's the only member.\n",
    "img = (np.expand_dims(img,0))\n",
    "\n",
    "print(img.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "EQ5wLTkcCaXY"
   },
   "source": [
    "Now predict the image:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "o_rzNSdrCaXY"
   },
   "outputs": [],
   "source": [
    "predictions_single = model.predict(img)\n",
    "\n",
    "print(predictions_single)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "6Ai-cpLjO-3A"
   },
   "outputs": [],
   "source": [
    "plot_value_array(0, predictions_single, test_labels)\n",
    "plt.xticks(range(10), class_names, rotation=45)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "cU1Y2OAMCaXb"
   },
   "source": [
    "`model.predict` returns a list of lists, one for each image in the batch of data. Grab the predictions for our (only) image in the batch:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "2tRmdq_8CaXb"
   },
   "outputs": [],
   "source": [
    "prediction_result = np.argmax(predictions_single[0])\n",
    "print(prediction_result)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "YFc2HbEVCaXd"
   },
   "source": [
    "And, as before, the model predicts a label of 9."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Classify your own Data\n",
    "\n",
    "Produce your own 28 x 28 pixel grayscale image of an article of clothing, upload it to here, and classify. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[[209 135  71 255]\n",
      "  [216 145  87 255]\n",
      "  [218 152  84 255]\n",
      "  ...\n",
      "  [223 173 111 255]\n",
      "  [208 134  69 255]\n",
      "  [198 107  40 255]]\n",
      "\n",
      " [[207 161  97 255]\n",
      "  [220 178 133 255]\n",
      "  [226 191 139 255]\n",
      "  ...\n",
      "  [230 191 126 255]\n",
      "  [212 144  71 255]\n",
      "  [198 107  35 255]]\n",
      "\n",
      " [[202 178 121 255]\n",
      "  [213 194 157 255]\n",
      "  [221 206 159 255]\n",
      "  ...\n",
      "  [227 196 128 255]\n",
      "  [210 150  75 255]\n",
      "  [193 108  37 255]]\n",
      "\n",
      " ...\n",
      "\n",
      " [[207 173 113 255]\n",
      "  [205 174 120 255]\n",
      "  [204 182 122 255]\n",
      "  ...\n",
      "  [184 122  88 255]\n",
      "  [187 109  79 255]\n",
      "  [196 107  74 255]]\n",
      "\n",
      " [[204 169 113 255]\n",
      "  [201 174 122 255]\n",
      "  [202 184 131 255]\n",
      "  ...\n",
      "  [201 152 115 255]\n",
      "  [210 147 109 255]\n",
      "  [210 130  86 255]]\n",
      "\n",
      " [[206 172 104 255]\n",
      "  [203 176 111 255]\n",
      "  [202 181 118 255]\n",
      "  ...\n",
      "  [210 164 107 255]\n",
      "  [222 165 103 255]\n",
      "  [221 150  88 255]]]\n",
      "[[135 145 152 ... 173 134 107]\n",
      " [161 178 191 ... 191 144 107]\n",
      " [178 194 206 ... 196 150 108]\n",
      " ...\n",
      " [173 174 182 ... 122 109 107]\n",
      " [169 174 184 ... 152 147 130]\n",
      " [172 176 181 ... 164 165 150]]\n"
     ]
    }
   ],
   "source": [
    "# 1. With OpenCV\n",
    "\n",
    "# import cv2\n",
    "# img = cv2.imread(\"burgerfries.png\")\n",
    "# print(img)\n",
    "\n",
    "# 2. With SciPY\n",
    "\n",
    "# from scipy import misc\n",
    "# img = misc.imread(\"burgerfries.png\")\n",
    "# print(img)\n",
    "\n",
    "# 3. With Python Imaging Library (PIL)\n",
    "\n",
    "from PIL import Image\n",
    "Image.open(\"burgerfries.png\")\n",
    "print(img)\n",
    "img = img[...,1]\n",
    "print(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "basic_classification.ipynb",
   "private_outputs": true,
   "provenance": [],
   "toc_visible": true,
   "version": "0.3.2"
  },
  "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}