07- Convolutional Neural Networks/Lab/Lab07-MRI_Dementia.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Machine Learning for Neuroscience, <br>Department of Brain Sciences, Faculty of Medicine, <br> Imperial College London\n",
    "### Contributors: Francesca Palermo, Nan Fletcher-Lloyd, Alex Capstick, Yu Chen\n",
    "**Winter 2022**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Convolutional Neural Network (ConvNet/CNN)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Convolutional Neural Networks (ConvNet/CNN) are Deep Learning models that can take an input image, assign relevance (learnable weights and biases) to various aspects/objects in the image, and distinguish between them. <br />\n",
    "\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"./_dependents/rgb.png?raw=true\" width=\"500\" /></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You often see images divided into three colour planes — Red, Green, and Blue — RGB. <br />\n",
    "But they can be stored in a variety of colour spaces, including Grayscale, RGB, HSV, CMYK, and others. <br />\n",
    "\n",
    "\n",
    "With the basic machine learning techniques, it is required to hand-engineering features and analyse the figures dimension by dimension. \n",
    "Imagine now taking a photo which exceeds 8K (7680x4320) dimensions and how computationally intensive it will be.\n",
    "The ConvNet's job is to compress the images into a format that is easier to process while preserving elements that are important for obtaining a decent prediction. This is critical for designing an architecture that is capable of learning features while also being scalable to large datasets. <br />\n",
    "So, while basic machine learning techniques require hand-engineering of features, ConvNets can learn these filters/characteristics through training.\n",
    "\n",
    "\n",
    "Below an example of a CNN architecture for classification. Starting from a picture, it analyses through its layers and output the correct class of object present in the image (a car in this case)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"./_dependents/basic_architecture.png?raw=true\" width=\"800\" /></center>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using PyTorch for creating a CNN architecture"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this experiment we will use the ```PyTorch``` library for creating our own CNN model.\n",
    "\n",
    "First of all, let us import the necessary libraries with the following commands:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "d:\\Anaconda3\\envs\\imperial\\lib\\site-packages\\tqdm\\auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "import torch # PyTorch is an open source machine learning framework.\n",
    "import torchvision # The torchvision package consists of popular datasets, model architectures, and common image transformations for computer vision.\n",
    "from torchvision import transforms\n",
    "from PIL import Image # PIL is a library to process images\n",
    "import numpy as np # package to work with numbers\n",
    "from sklearn.metrics import recall_score, precision_score, f1_score, accuracy_score, confusion_matrix\n",
    "import os\n",
    "import glob\n",
    "from pandas.core.common import flatten\n",
    "import random\n",
    "from collections import Counter\n",
    "from torch.utils.data import Dataset\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import train_test_split\n",
    "import cv2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can create a ```device``` variable to check if a GPU is available.\n",
    "If the output of device variable is 'cuda', then it means that PyTorch will use a GPU for the analysis. Otherwise, CPU is used instead."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "GPU available: True\n",
      "NVIDIA GeForce RTX 3070\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "device(type='cuda')"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Device configuration\n",
    "print('GPU available: {}' .format(torch.cuda.is_available())) \n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "train_on_gpu = torch.cuda.is_available()\n",
    "device\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load MRI Image Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will now load the  ```Alzheimer MRI Preprocessed Dataset``` dataset that we will use for the laboratory. For easier access, the dataset has already been downloaded and need to be unzip in a folder of your choice which will be called the ```dataset_path```. This dataset is made of around 6400 images of 128 x 128 pixels. For our analysis we have split the dataset into two classes: Dementia and Healthy.\n",
    "\n",
    "\n",
    "Additional information on the dataset can be found [here](https://www.kaggle.com/datasets/sachinkumar413/alzheimer-mri-dataset)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We need to first unzip the archive file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import shutil\n",
    "shutil.unpack_archive('./database/MRI_database.zip', './database/') # specificy the file to unpack and where to unpack it"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can then start load the data and creating the training, testing and validation datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "classes: ['Dementia' 'Healthy']\n",
      "Counter({'Dementia': 3200, 'Healthy': 3200})\n",
      "Original dataset size: 6400\n",
      "\n",
      "Split datasets size:\n",
      "Train-set 4608 Counter({'Healthy': 2311, 'Dementia': 2297})\n",
      "Validation-set 1152 Counter({'Healthy': 590, 'Dementia': 562})\n",
      "Test-set 640 Counter({'Dementia': 341, 'Healthy': 299})\n"
     ]
    }
   ],
   "source": [
    "from torchvision import datasets\n",
    "from torchvision.transforms import ToTensor\n",
    "from torch.utils.data.sampler import SubsetRandomSampler\n",
    "\n",
    "# convert data to a normalized torch.FloatTensor\n",
    "transform = transforms.Compose([\n",
    "    transforms.ToTensor(),\n",
    "    # transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))\n",
    "    ])\n",
    "\n",
    "image_paths = [] #to store image paths in list\n",
    "classes = [] #to store class values\n",
    "dataset_path = './database/'\n",
    "for subdir, dirs, files in os.walk(dataset_path):\n",
    "    for file in files:\n",
    "        if file.endswith('.jpg'):\n",
    "            classes.append(subdir.split('/')[-1].split('\\\\')[-1]) \n",
    "            image_paths.append(os.path.join(subdir, file))\n",
    "    \n",
    "image_paths = list(flatten(image_paths))\n",
    "\n",
    "print('classes: {}'.format(np.unique(classes))) \n",
    "print(Counter(classes))\n",
    "print(\"Original dataset size: {}\".format(len(image_paths)))\n",
    "\n",
    "# train 70% validation 20% test 10%\n",
    "X_train, X_test, y_train, y_test = train_test_split(image_paths, classes, test_size=0.1, random_state=42)\n",
    "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=0.2, random_state=42)\n",
    "n_classes = len(np.unique(y_train))\n",
    "\n",
    "print('\\nSplit datasets size:\\nTrain-set {} {}\\nValidation-set {} {}\\nTest-set {} {}'.format(len(X_train), Counter(y_train), len(X_val), Counter(y_val), len(X_test), Counter(y_test),))\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since the labels are categorically encoded, we can create a dictionary to associate each label to a different integer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Dementia': 0, 'Healthy': 1}"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "idx_to_class = {i:j for i, j in enumerate(np.unique(classes))}\n",
    "class_to_idx = {value:key for key,value in idx_to_class.items()}\n",
    "class_to_idx\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preparing data with DataLoader"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Dataset retrieves our dataset’s features and labels one sample at a time. While training a model, we want to pass samples in “minibatches”, reshuffle the data at every epoch to reduce model overfitting.\n",
    "\n",
    "```DataLoader``` is an iterable variable that we can use in PyTorch to automatically perform these functions.\n",
    "\n",
    "For additional information on ```DataLoader```, you can refer to the [official documentation](https://pytorch.org/docs/stable/data.html)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To load the dataset into pytorch, we need to create a class which will be able to correctly load each image in the correct set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "class MRIDataset(Dataset):\n",
    "    def __init__(self, image_paths, classes, transform=None):\n",
    "        self.image_paths = image_paths\n",
    "        self.transform = transform\n",
    "        self.classes = classes\n",
    "        \n",
    "    def __len__(self):\n",
    "        return len(self.image_paths)\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        image_filepath = self.image_paths[idx]\n",
    "        image = cv2.imread(image_filepath)\n",
    "        label = self.classes[idx]\n",
    "        label = class_to_idx[label]\n",
    "        # label = class_to_idx[label]\n",
    "        if self.transform is not None:\n",
    "            image = self.transform(image)\n",
    "            \n",
    "        return image, label"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can then create the training, validation and testing sets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dataset = MRIDataset(X_train, y_train, transform)\n",
    "val_dataset = MRIDataset(X_val, y_val, transform)\n",
    "test_dataset = MRIDataset(X_test, y_test, transform)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below we will load the dataset into DataLoader of batch_size=64 features and labels respectively. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from torch.utils.data import DataLoader\n",
    "batch_size = 64\n",
    "\n",
    "train_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True)\n",
    "valid_loader = DataLoader(val_dataset, batch_size=batch_size, shuffle=True)\n",
    "test_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can iterate over the DataLoader variable and extracting the features of the image and the corresponding label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3, 128, 128)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x576 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def imshow(img):\n",
    "    img = img / 2 + 0.5  # unnormalize\n",
    "    # plt.imshow()  # convert from Tensor image\n",
    "\n",
    "# obtain one batch of training images\n",
    "dataiter = iter(test_loader)\n",
    "images, labels = dataiter.next()  # if this line gives you error, change it into: images, labels = next(dataiter)\n",
    "images = images.numpy() # convert images to numpy for display\n",
    "ncols = 5\n",
    "nrows = 2\n",
    "figure, ax = plt.subplots(nrows=nrows, ncols=ncols, figsize=(12, 8))\n",
    "for i in range(nrows*ncols):\n",
    "    ax.ravel()[i].imshow((np.transpose(images[i], (1, 2, 0))* 255).astype(np.uint8))\n",
    "    ax.ravel()[i].set_axis_off()\n",
    "    ax.ravel()[i].set_title(idx_to_class[labels[i].item()])\n",
    "images[i].shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create CNN Architecture"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will now create the architecture for the CNN. In this example we will use two convolutional layers followed by three fully-connected layers. \n",
    "\n",
    "When creating the network, we also need to specify the activation function of each layer. We will use the Rectified Linear Units (ReLU) in this case.\n",
    "\n",
    "If necessary, we can also introduce a regularisation layer to reduce possible overfitting. A simple way to do this can be to use dropout layers.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import torch.nn.functional as F\n",
    "\n",
    "# define the CNN architecture\n",
    "\n",
    "class CNN(nn.Module):\n",
    "    def __init__(self, n_class):\n",
    "        super(CNN, self).__init__()\n",
    "               \n",
    "        self.conv1 = nn.Conv2d(3, 16, 3) #n_channels = 3, rgb image. These parameters can be modified and optimised based on the task and/or dataset\n",
    "        self.batchnorm1 = nn.BatchNorm2d(16)\n",
    "        self.pool = nn.MaxPool2d(2, 2) # Applies 2D max-pooling to reduce the spatial dimensions of the input volume\n",
    "        self.conv2 = nn.Conv2d(16, 32, 3)  \n",
    "        self.batchnorm2 = nn.BatchNorm2d(32)\n",
    "        self.conv2_drop = nn.Dropout2d(0.2)\n",
    "        self.conv3 = nn.Conv2d(32, 64, 3)  \n",
    "        self.batchnorm3 = nn.BatchNorm2d(64)\n",
    "        self.conv3_drop = nn.Dropout2d(0.25)\n",
    "        self.fc1 = nn.Linear(12544, 128) # Fully connected layers\n",
    "        self.fc2 = nn.Linear(128, 64) \n",
    "        self.fc3 = nn.Linear(64, n_class)\n",
    "        self.flatten = nn.Flatten()\n",
    "    \n",
    "    \n",
    "    # The forward pass defines the way we compute the output using the given layers and functions.\n",
    "    def forward(self, x):\n",
    "        x = self.pool(F.relu(self.conv1(x))) # activation function\n",
    "        x = self.batchnorm1(x)\n",
    "        x = self.pool(F.relu(self.conv2(x)))  \n",
    "        x = self.batchnorm2(x)\n",
    "        x = self.conv2_drop(x)\n",
    "        x = self.pool(F.relu(self.conv3(x))) \n",
    "        x = self.batchnorm3(x)\n",
    "        x = self.conv3_drop(x)\n",
    "        x = self.flatten(x)\n",
    "        x = F.relu(self.fc1(x))\n",
    "        x = F.relu(self.fc2(x))\n",
    "        x = self.fc3(x)\n",
    "        return x\n",
    "\n",
    "# create the model CNN\n",
    "model = CNN(n_classes)\n",
    "\n",
    "# move tensors to GPU if CUDA is available\n",
    "if train_on_gpu:\n",
    "    model.to(device=device)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can print the architecture of the CNN network to investigate its layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "----------------------------------------------------------------\n",
      "        Layer (type)               Output Shape         Param #\n",
      "================================================================\n",
      "            Conv2d-1         [-1, 16, 126, 126]             448\n",
      "         MaxPool2d-2           [-1, 16, 63, 63]               0\n",
      "       BatchNorm2d-3           [-1, 16, 63, 63]              32\n",
      "            Conv2d-4           [-1, 32, 61, 61]           4,640\n",
      "         MaxPool2d-5           [-1, 32, 30, 30]               0\n",
      "       BatchNorm2d-6           [-1, 32, 30, 30]              64\n",
      "         Dropout2d-7           [-1, 32, 30, 30]               0\n",
      "            Conv2d-8           [-1, 64, 28, 28]          18,496\n",
      "         MaxPool2d-9           [-1, 64, 14, 14]               0\n",
      "      BatchNorm2d-10           [-1, 64, 14, 14]             128\n",
      "        Dropout2d-11           [-1, 64, 14, 14]               0\n",
      "          Flatten-12                [-1, 12544]               0\n",
      "           Linear-13                  [-1, 128]       1,605,760\n",
      "           Linear-14                   [-1, 64]           8,256\n",
      "           Linear-15                    [-1, 2]             130\n",
      "================================================================\n",
      "Total params: 1,637,954\n",
      "Trainable params: 1,637,954\n",
      "Non-trainable params: 0\n",
      "----------------------------------------------------------------\n",
      "Input size (MB): 0.19\n",
      "Forward/backward pass size (MB): 5.24\n",
      "Params size (MB): 6.25\n",
      "Estimated Total Size (MB): 11.68\n",
      "----------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "from torchsummary import summary\n",
    "\n",
    "summary(model, (3, 128, 128))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define Loss and Optimser "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Loss**\n",
    "\n",
    "Deep neural network are trained via an algorithm called stochastic gradient descent optimization.\n",
    "It is necessary to repeatedly estimate the error for the model's current state as part of the optimiaation method. In order to update the weights and lower the loss on the subsequent evaluation, it is necessary to select an error function, also known as a loss function, that may be used to estimate the loss of the model.\n",
    "\n",
    "There exist multiple and different losses based on the scope of the analysis.\n",
    "In this experiment, we are interested in classifying the number present in the image. We are then performing a multi-class classification. Our goal is to predict the probability of the sample belonging to each known class (from number 0 to 9).\n",
    "\n",
    "We can then use a loss called ```Cross-entropy```. It will calculate a score to summarise the average difference between the actual and predicted probability distributions for all classes. To achieve the best results we have to minimise this loss\n",
    "\n",
    "\n",
    "\n",
    "**Optimiser**\n",
    "\n",
    "When training the model, we need to minimise the loss function and update the weights of the model at each epoch. \n",
    "In other words, we need to optimise the neural network.\n",
    "An optimiser is a function or an algorithm that modifies the attributes of the neural network, such as weights and learning rate. It works for reducing the loss and improve the overall accuracy.\n",
    "\n",
    "There are various types of optimiser. In our case, we will use ```Adam``` which is one of the most popular gradient descent optimisation algorithms.\n",
    "\n",
    "When setting up the optimiser, we also need to decide at which speed the model will learn. This is done via the learning rate.\n",
    "It will tell the model how fast or slow we will move towards the optimal weights\n",
    "<center><img src=\"./_dependents/learning_rate.png\" width=\"300\" /></center>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Adam (\n",
       "Parameter Group 0\n",
       "    amsgrad: False\n",
       "    betas: (0.9, 0.999)\n",
       "    capturable: False\n",
       "    eps: 1e-08\n",
       "    foreach: None\n",
       "    lr: 0.001\n",
       "    maximize: False\n",
       "    weight_decay: 0\n",
       ")"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "criterion = torch.nn.CrossEntropyLoss().to(device)  # cross-entropy loss for classification\n",
    "optimiser = torch.optim.Adam(model.parameters(), lr=0.001) \n",
    "\n",
    "optimiser"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train the model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have finally set up all the parameters and we can move on to train the model on our dataset.\n",
    "We need to set the number of iterations that we want to train our model. This number needs to be carefully decided: if it is too low, then we risk of underfitting our model, if it is too high, we risk of overfitting it.\n",
    "<center><img src=\"./_dependents/overfitting_model.png\" width=\"300\" /></center>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[INFO] TRAINING EPOCH 1\n",
      "[INFO] Epoch 1, Training Loss: 0.594679\n",
      "[INFO] VALIDATION EPOCH 1\n",
      "[INFO] Epoch 1, Validation Loss: 0.566541\n",
      "Validation loss decreased (inf --> 0.566541).  Saving model ...\n",
      "-----------------------------------------------------------\n",
      "[INFO] TRAINING EPOCH 2\n",
      "[INFO] Epoch 2, Training Loss: 0.470033\n",
      "[INFO] VALIDATION EPOCH 2\n",
      "[INFO] Epoch 2, Validation Loss: 0.404361\n",
      "Validation loss decreased (0.566541 --> 0.404361).  Saving model ...\n",
      "-----------------------------------------------------------\n",
      "[INFO] TRAINING EPOCH 3\n",
      "[INFO] Epoch 3, Training Loss: 0.351587\n",
      "[INFO] VALIDATION EPOCH 3\n",
      "[INFO] Epoch 3, Validation Loss: 0.293041\n",
      "Validation loss decreased (0.404361 --> 0.293041).  Saving model ...\n",
      "-----------------------------------------------------------\n",
      "[INFO] TRAINING EPOCH 4\n",
      "[INFO] Epoch 4, Training Loss: 0.219788\n",
      "[INFO] VALIDATION EPOCH 4\n",
      "[INFO] Epoch 4, Validation Loss: 0.415056\n",
      "-----------------------------------------------------------\n",
      "[INFO] TRAINING EPOCH 5\n",
      "[INFO] Epoch 5, Training Loss: 0.156559\n",
      "[INFO] VALIDATION EPOCH 5\n",
      "[INFO] Epoch 5, Validation Loss: 0.134685\n",
      "Validation loss decreased (0.293041 --> 0.134685).  Saving model ...\n",
      "-----------------------------------------------------------\n",
      "[INFO] TRAINING EPOCH 6\n",
      "[INFO] Epoch 6, Training Loss: 0.087298\n",
      "[INFO] VALIDATION EPOCH 6\n",
      "[INFO] Epoch 6, Validation Loss: 0.404854\n",
      "-----------------------------------------------------------\n",
      "[INFO] TRAINING EPOCH 7\n",
      "[INFO] Epoch 7, Training Loss: 0.066246\n",
      "[INFO] VALIDATION EPOCH 7\n",
      "[INFO] Epoch 7, Validation Loss: 0.071785\n",
      "Validation loss decreased (0.134685 --> 0.071785).  Saving model ...\n",
      "-----------------------------------------------------------\n",
      "[INFO] TRAINING EPOCH 8\n",
      "[INFO] Epoch 8, Training Loss: 0.042099\n",
      "[INFO] VALIDATION EPOCH 8\n",
      "[INFO] Epoch 8, Validation Loss: 0.092898\n",
      "-----------------------------------------------------------\n",
      "[INFO] TRAINING EPOCH 9\n",
      "[INFO] Epoch 9, Training Loss: 0.041123\n",
      "[INFO] VALIDATION EPOCH 9\n",
      "[INFO] Epoch 9, Validation Loss: 0.076863\n",
      "-----------------------------------------------------------\n",
      "[INFO] TRAINING EPOCH 10\n",
      "[INFO] Epoch 10, Training Loss: 0.026266\n",
      "[INFO] VALIDATION EPOCH 10\n",
      "[INFO] Epoch 10, Validation Loss: 0.105939\n",
      "-----------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = CNN(n_classes)\n",
    "if train_on_gpu:\n",
    "    model.to(device)\n",
    "\n",
    "criterion = torch.nn.CrossEntropyLoss().to(device)  # cross-entropy loss for classification\n",
    "optimiser = torch.optim.Adam(model.parameters(), lr=0.001) \n",
    "\n",
    "# number of epochs to train the model\n",
    "n_epochs = 10\n",
    "\n",
    "#List to store loss to visualize\n",
    "train_losslist = []\n",
    "valid_loss_min = np.Inf # track change in validation loss\n",
    "\n",
    "for epoch in range(1, n_epochs+1):\n",
    "\n",
    "    # keep track of training and validation loss\n",
    "    train_loss = 0.0\n",
    "    valid_loss = 0.0\n",
    "    \n",
    "    ###################\n",
    "    # train the model #\n",
    "    ###################\n",
    "    print('[INFO] TRAINING EPOCH {}'.format(epoch))\n",
    "    model.train()\n",
    "    for data, target in train_loader:\n",
    "        # move tensors to GPU if CUDA is available\n",
    "        if train_on_gpu:\n",
    "            data, target = data.to(device=device), target.to(device=device)\n",
    "        # clear the gradients of all optimized variables\n",
    "        optimiser.zero_grad()\n",
    "        # forward pass: compute predicted outputs by passing inputs to the model\n",
    "        output = model(data)\n",
    "        # calculate the batch loss\n",
    "        loss = criterion(output, target)\n",
    "        # backward pass: compute gradient of the loss with respect to model parameters\n",
    "        loss.backward()\n",
    "        # perform a single optimization step (parameter update)\n",
    "        optimiser.step()\n",
    "        # update training loss\n",
    "        train_loss += loss.item()*data.size(0)\n",
    "    \n",
    "    train_loss = train_loss/len(train_loader.dataset)\n",
    "    train_losslist.append(train_loss)\n",
    "    print('[INFO] Epoch {}, Training Loss: {:.6f}'.format(epoch, train_loss))\n",
    "        \n",
    "    ######################    \n",
    "    # validate the model #\n",
    "    ######################\n",
    "    print('[INFO] VALIDATION EPOCH {}'.format(epoch))\n",
    "    model.eval()\n",
    "    for data, target in valid_loader:\n",
    "        # move tensors to GPU if CUDA is available\n",
    "        if train_on_gpu:\n",
    "            data, target = data.to(device=device), target.to(device=device)\n",
    "        # forward pass: compute predicted outputs by passing inputs to the model\n",
    "        output = model(data)\n",
    "        # calculate the batch loss\n",
    "        loss = criterion(output, target)\n",
    "        # update average validation loss \n",
    "        valid_loss += loss.item()*data.size(0)\n",
    "    \n",
    "    # calculate average losses\n",
    "    valid_loss = valid_loss/len(valid_loader.dataset)\n",
    "    print('[INFO] Epoch {}, Validation Loss: {:.6f}'.format(epoch, valid_loss))\n",
    "\n",
    "    # save model if validation loss has decreased\n",
    "    if valid_loss <= valid_loss_min:\n",
    "        print('Validation loss decreased ({:.6f} --> {:.6f}).  Saving model ...'.format(\n",
    "        valid_loss_min,\n",
    "        valid_loss))\n",
    "        torch.save(model.state_dict(), 'model_MRI.pt')\n",
    "        valid_loss_min = valid_loss\n",
    "    print('-----------------------------------------------------------')    \n",
    "\n",
    "plt.plot([*range(n_epochs)], train_losslist)\n",
    "plt.xlabel(\"Epoch\")\n",
    "plt.ylabel(\"Loss\")\n",
    "plt.title(\"Performance of CNN model\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test the model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will now load the model which achieved the best results during training and we will test it on the testing set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.load_state_dict(torch.load('model_MRI.pt'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Loss: 0.095937\n",
      "\n",
      "Test Accuracy of Dementia: 95% (326/341)\n",
      "Test Accuracy of Healthy: 98% (294/299)\n",
      "\n",
      "Test Accuracy (Overall): 96% (620/640)\n"
     ]
    }
   ],
   "source": [
    "# track test loss\n",
    "test_loss = 0.0\n",
    "class_correct = list(0. for i in range(10))\n",
    "class_total = list(0. for i in range(10))\n",
    "\n",
    "total_test_predicted_labels = []\n",
    "total_test_true_labels = []\n",
    "\n",
    "model.eval()\n",
    "# iterate over test data\n",
    "for data, target in test_loader:\n",
    "    # move tensors to GPU if CUDA is available\n",
    "    if train_on_gpu:\n",
    "        data, target = data.to(device=device), target.to(device=device)\n",
    "    # forward pass: compute predicted outputs by passing inputs to the model\n",
    "    output = model(data)\n",
    "    # calculate the batch loss\n",
    "    loss = criterion(output, target)\n",
    "    # update test loss \n",
    "    test_loss += loss.item()*data.size(0)\n",
    "    # convert output probabilities to predicted class\n",
    "    _, pred = torch.max(output, 1)    \n",
    "    # compare predictions to true label\n",
    "    correct_tensor = pred.eq(target.data.view_as(pred))\n",
    "    correct = np.squeeze(correct_tensor.numpy()) if not train_on_gpu else np.squeeze(correct_tensor.cpu().numpy())\n",
    "    \n",
    "    total_test_true_labels.append(target.cpu().numpy())\n",
    "    total_test_predicted_labels.append(pred.cpu().numpy())\n",
    "\n",
    "    # calculate test accuracy for each object class\n",
    "    for i in range(target.shape[0]):\n",
    "        label = target.data[i]\n",
    "        class_correct[label] += correct[i].item()\n",
    "        class_total[label] += 1\n",
    "        \n",
    "        \n",
    "y_true = np.asarray([item for sublist in total_test_true_labels for item\n",
    "                             in sublist])\n",
    "y_pred = np.asarray([item for sublist in total_test_predicted_labels for\n",
    "                               item in sublist])        \n",
    "\n",
    "# average test loss\n",
    "test_loss = test_loss/len(test_loader.dataset)\n",
    "print('Test Loss: {:.6f}\\n'.format(test_loss))\n",
    "\n",
    "for i in range(2):\n",
    "    if class_total[i] > 0:\n",
    "        print('Test Accuracy of %5s: %2d%% (%2d/%2d)' % (\n",
    "            np.unique(classes)[i], 100 * class_correct[i] / class_total[i],\n",
    "            np.sum(class_correct[i]), np.sum(class_total[i])))\n",
    "    else:\n",
    "        print('Test Accuracy of %5s: N/A (no training examples)' % (classes[i]))\n",
    "\n",
    "print('\\nTest Accuracy (Overall): %2d%% (%2d/%2d)' % (\n",
    "    100. * np.sum(class_correct) / np.sum(class_total),\n",
    "    np.sum(class_correct), np.sum(class_total)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Investigate results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ACCURACY: 0.97\n",
      " RECALL: 0.97\n",
      " PRECISION: 0.97\n",
      " F1: 0.97\n",
      "  \n",
      "[[326  15]\n",
      " [  5 294]]\n"
     ]
    }
   ],
   "source": [
    "accuracy = accuracy_score(y_true, y_pred)\n",
    "recall = recall_score(y_true, y_pred, average='weighted')\n",
    "precision = precision_score(y_true, y_pred, average='weighted')\n",
    "f1 = f1_score(y_true, y_pred, average='weighted')\n",
    "cf = confusion_matrix(y_true, y_pred)\n",
    "print('ACCURACY: {:.2f}\\n RECALL: {:.2f}\\n PRECISION: {:.2f}\\n F1: {:.2f}\\n  '.format(accuracy, \n",
    "                                                                                      recall,\n",
    "                                                                                      precision, \n",
    "                                                                                      f1))\n",
    "print(cf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 360x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "fig, ax = plt.subplots(figsize=(5,5))\n",
    "sns.heatmap(cf, annot=True, ax = ax, cmap='Blues', fmt='g')\n",
    "ax.set_xticklabels(np.unique(classes))\n",
    "ax.set_yticklabels(np.unique(classes))\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create Train and Test Methods "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can further analyse the model by changing various values in the architecture of the network. \n",
    "Instead of having training and testing in the script, we can create two functions which will be called whenever we need to train or test the model. We can load the train and test methods from Lecture 6 - Neural Networks."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from utils import module_from_file\n",
    "train_model = module_from_file(\n",
    "    \"model_trainer\",\n",
    "    \"../../Topic-06- Neural Networks/Lab/model_trainer.py\"\n",
    "    ).train\n",
    "\n",
    "predict_model = module_from_file(\n",
    "    \"model_predictor\",\n",
    "    \"../../Topic-06- Neural Networks/Lab/model_predictor.py\"\n",
    "    ).predict"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After importing the methods, we can start the training of the model by passing the train and val loader, the number of epochs, the optimiser and criterion."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Training: 100%|██████████| 10/10 [00:40<00:00,  4.09s/it, Train Loss=0.025, Val Loss=0.200]\n"
     ]
    }
   ],
   "source": [
    "model = CNN(n_classes)\n",
    "if train_on_gpu:\n",
    "    model.to(device)\n",
    "\n",
    "criterion = torch.nn.CrossEntropyLoss().to(device)  # cross-entropy loss for classification\n",
    "optimiser = torch.optim.Adam(model.parameters(), lr=0.001) \n",
    "\n",
    "\n",
    "model, (train_loss_dict, val_loss_dict) = train_model(\n",
    "    model=model,\n",
    "    train_loader=train_loader,\n",
    "    val_loader=valid_loader,\n",
    "    n_epochs=n_epochs,\n",
    "    optimiser=optimiser,\n",
    "    criterion=criterion,\n",
    "    )\n",
    "# torch.save(model.state_dict(), 'model_MRI.pt') # if you want to save the model, you can change the name as you prefer"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Once the training is finished, we can then test the model on the test dataset and predict the possible labels. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Predicting: 100%|██████████| 10/10 [00:00<00:00, 16.30it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 0 1 1 0]\n",
      "[0 0 1 1 0]\n"
     ]
    }
   ],
   "source": [
    "y_pred, y_true = predict_model(model, test_loader=test_loader)\n",
    "# finding the class of max probability\n",
    "_, y_pred = torch.max(y_pred, dim=1) \n",
    "y_pred = y_pred.numpy()\n",
    "y_true = y_true.numpy()\n",
    "print(y_pred[0:5])\n",
    "print(y_true[0:5])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can then investigate the results as done in the previous section"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# simple methods to calculate all the required metrics for analysing the model. You can also return the metrics, if you want to further analyse them\n",
    "def calculate_metrics(y_true, y_pred):\n",
    "    accuracy = accuracy_score(y_true, y_pred)\n",
    "    recall = recall_score(y_true, y_pred, average='weighted')\n",
    "    precision = precision_score(y_true, y_pred, average='weighted')\n",
    "    f1 = f1_score(y_true, y_pred, average='weighted')\n",
    "    cf = confusion_matrix(y_true, y_pred)\n",
    "    print('ACCURACY: {:.2f}\\n RECALL: {:.2f}\\n PRECISION: {:.2f}\\n F1: {:.2f}\\n  '.format(accuracy, \n",
    "                                                                                        recall,\n",
    "                                                                                        precision, \n",
    "                                                                                        f1))\n",
    "    print(cf)\n",
    "    return accuracy, recall, precision, f1, cf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ACCURACY: 0.94\n",
      " RECALL: 0.94\n",
      " PRECISION: 0.95\n",
      " F1: 0.94\n",
      "  \n",
      "[[303  38]\n",
      " [  0 299]]\n"
     ]
    }
   ],
   "source": [
    "original_accuracy, original_recall, original_precision, original_f1, original_cf = calculate_metrics(y_true, y_pred)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Additional Questions"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1) We have introduced a CNN model and two methods for training and making predictions on a dataset. We can now investigate what would happen when using the same CNN model as before but changing the learning rate in the optimiser. Try training with different learning rates and investigate if there are any changes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Training: 100%|██████████| 10/10 [00:39<00:00,  3.96s/it, Train Loss=0.696, Val Loss=0.706]\n",
      "Predicting: 100%|██████████| 10/10 [00:00<00:00, 26.43it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ACCURACY: 0.53\n",
      " RECALL: 0.53\n",
      " PRECISION: 0.28\n",
      " F1: 0.37\n",
      "  \n",
      "[[341   0]\n",
      " [299   0]]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "d:\\Anaconda3\\envs\\imperial\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1334: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    }
   ],
   "source": [
    "n_epochs = 10 \n",
    "\n",
    "model = CNN(n_classes)\n",
    "\n",
    "criterion = torch.nn.CrossEntropyLoss().to(device)  # cross-entropy loss for classification\n",
    "optimiser = torch.optim.Adam(model.parameters(), lr=0.1) \n",
    "\n",
    "model, (lr_train_loss_dict, lr_val_loss_dict) = train_model(\n",
    "    model=model,\n",
    "    train_loader=train_loader,\n",
    "    val_loader=valid_loader,\n",
    "    n_epochs=n_epochs,\n",
    "    optimiser=optimiser,\n",
    "    criterion=criterion,\n",
    "    )\n",
    "    \n",
    "y_pred, y_true = predict_model(model, test_loader=test_loader)\n",
    "# finding the class of max probability\n",
    "_, y_pred = torch.max(y_pred, dim=1) \n",
    "y_pred = y_pred.numpy()\n",
    "y_true = y_true.numpy()\n",
    "lr_accuracy, lr_recall, lr_precision, lr_f1, lr_cf = calculate_metrics(y_true, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Performance of CNN model')"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "# plot training loss\n",
    "ax.plot(\n",
    "    lr_train_loss_dict['step'], lr_train_loss_dict['loss'],\n",
    "    label='Train'\n",
    "    )\n",
    "\n",
    "# plot training loss\n",
    "ax.plot(\n",
    "    lr_val_loss_dict['step'], lr_val_loss_dict['loss'],\n",
    "    label='Val'\n",
    "    )\n",
    "ax.set_xlabel(\"Epoch\")\n",
    "ax.set_ylabel(\"Loss\")\n",
    "ax.set_title(\"Performance of CNN model\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2) Let us investigate now what would happen when changing the type of optimiser used for the classification. We can change the optimiser based on our need and task. <br>\n",
    "More information can be found: https://pytorch.org/tutorials/beginner/basics/optimization_tutorial.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Training: 100%|██████████| 10/10 [00:38<00:00,  3.84s/it, Train Loss=0.546, Val Loss=0.549]\n",
      "Predicting: 100%|██████████| 10/10 [00:00<00:00, 25.60it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ACCURACY: 0.75\n",
      " RECALL: 0.75\n",
      " PRECISION: 0.76\n",
      " F1: 0.75\n",
      "  \n",
      "[[245  96]\n",
      " [ 61 238]]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "n_epochs = 10 \n",
    "\n",
    "model = CNN(n_classes)\n",
    "\n",
    "criterion = torch.nn.CrossEntropyLoss().to(device)  # cross-entropy loss for classification\n",
    "optimiser = torch.optim.SGD(model.parameters(), lr=0.001)\n",
    "\n",
    "model, (opt_train_loss_dict, opt_val_loss_dict) = train_model(\n",
    "    model=model,\n",
    "    train_loader=train_loader,\n",
    "    val_loader=valid_loader,\n",
    "    n_epochs=n_epochs,\n",
    "    optimiser=optimiser,\n",
    "    criterion=criterion,\n",
    "    )\n",
    "    \n",
    "y_pred, y_true = predict_model(model, test_loader=test_loader)\n",
    "# finding the class of max probability\n",
    "_, y_pred = torch.max(y_pred, dim=1) \n",
    "y_pred = y_pred.numpy()\n",
    "y_true = y_true.numpy()\n",
    "opt_accuracy, opt_recall, opt_precision, opt_f1, opt_cf = calculate_metrics(y_true, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Performance of CNN model')"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "# plot training loss\n",
    "ax.plot(\n",
    "    opt_train_loss_dict['step'], opt_train_loss_dict['loss'],\n",
    "    label='Train'\n",
    "    )\n",
    "\n",
    "# plot training loss\n",
    "ax.plot(\n",
    "    opt_val_loss_dict['step'], opt_val_loss_dict['loss'],\n",
    "    label='Val'\n",
    "    )\n",
    "ax.set_xlabel(\"Epoch\")\n",
    "ax.set_ylabel(\"Loss\")\n",
    "ax.set_title(\"Performance of CNN model\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3) We can also change the number of epochs and check how the model learns depending on this parameter."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Training: 100%|██████████| 50/50 [03:09<00:00,  3.79s/it, Train Loss=0.008, Val Loss=0.041]\n",
      "Predicting: 100%|██████████| 10/10 [00:00<00:00, 28.37it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ACCURACY: 0.99\n",
      " RECALL: 0.99\n",
      " PRECISION: 0.99\n",
      " F1: 0.99\n",
      "  \n",
      "[[339   2]\n",
      " [  4 295]]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "n_epochs = 50\n",
    "\n",
    "model = CNN(n_classes)\n",
    "\n",
    "criterion = torch.nn.CrossEntropyLoss().to(device)  # cross-entropy loss for classification\n",
    "optimiser = torch.optim.Adam(model.parameters(), lr=0.0001) \n",
    "\n",
    "model, (epcs_train_loss_dict, epcs_val_loss_dict) = train_model(\n",
    "    model=model,\n",
    "    train_loader=train_loader,\n",
    "    val_loader=valid_loader,\n",
    "    n_epochs=n_epochs,\n",
    "    optimiser=optimiser,\n",
    "    criterion=criterion,\n",
    "    )\n",
    "    \n",
    "y_pred, y_true = predict_model(model, test_loader=test_loader)\n",
    "# finding the class of max probability\n",
    "_, y_pred = torch.max(y_pred, dim=1) \n",
    "y_pred = y_pred.numpy()\n",
    "y_true = y_true.numpy()\n",
    "epcs_accuracy, epcs_recall, epcs_precision, epcs_f1, epcs_cf = calculate_metrics(y_true, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Performance of CNN model')"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "# plot training loss\n",
    "ax.plot(\n",
    "    epcs_train_loss_dict['step'], epcs_train_loss_dict['loss'],\n",
    "    label='Train'\n",
    "    )\n",
    "\n",
    "# plot training loss\n",
    "ax.plot(\n",
    "    epcs_val_loss_dict['step'], epcs_val_loss_dict['loss'],\n",
    "    label='Val'\n",
    "    )\n",
    "ax.set_xlabel(\"Epoch\")\n",
    "ax.set_ylabel(\"Loss\")\n",
    "ax.set_title(\"Performance of CNN model\")\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4) Let's plot the above losses all together and compare them with the original loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1777193ff40>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "ax.plot(\n",
    "    lr_val_loss_dict['step'], lr_val_loss_dict['loss'],\n",
    "    label='Learning Rate'\n",
    "    )\n",
    "\n",
    "ax.plot(\n",
    "    opt_val_loss_dict['step'], opt_val_loss_dict['loss'],\n",
    "    label='Optimiser'\n",
    "    )\n",
    "\n",
    "ax.plot(\n",
    "    epcs_val_loss_dict['step'][0:25], epcs_val_loss_dict['loss'][0:25],\n",
    "    label='Epochs'\n",
    "    )\n",
    "\n",
    "ax.plot(\n",
    "    val_loss_dict['step'], val_loss_dict['loss'],\n",
    "    label='Original'\n",
    "    )\n",
    "\n",
    "\n",
    "ax.set_xlabel(\"Epoch\")\n",
    "ax.set_ylabel(\"Loss\")\n",
    "ax.set_title(\"Performance of CNN model\")\n",
    "plt.legend()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Additional Resources"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Documentation for the Convolutional Layer - https://pytorch.org/docs/stable/generated/torch.nn.Conv2d.html \n",
    "- Guide and discussion on optimisers - https://medium.com/mlearning-ai/optimizers-in-deep-learning-7bf81fed78a0\n",
    "- CIFAR10 Tutorial - https://pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html\n",
    "- CNN implementation on CIFAR10 - https://shonit2096.medium.com/cnn-on-cifar10-data-set-using-pytorch-34be87e09844"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.6 (default, Oct 18 2022, 12:41:40) \n[Clang 14.0.0 (clang-1400.0.29.202)]"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "31f2aee4e71d21fbe5cf8b01ff0e069b9275f58929596ceb00d14d90e3e16cd6"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}