shapTest.py

from __future__ import print_function
import argparse
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torchvision import datasets, transforms
from torch.optim.lr_scheduler import StepLR
import shap
import numpy as np
import matplotlib.pyplot as plt

# Define the CNN Model
class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 32, 3, 1)
        self.conv2 = nn.Conv2d(32, 64, 3, 1)
        self.dropout1 = nn.Dropout(0.25)
        self.dropout2 = nn.Dropout(0.5)
        self.fc1 = nn.Linear(9216, 128)
        self.fc2 = nn.Linear(128, 10)

    def forward(self, x):
        x = self.conv1(x)
        x = F.relu(x)
        x = self.conv2(x)
        x = F.relu(x)
        x = F.max_pool2d(x, 2)
        x = self.dropout1(x)
        x = torch.flatten(x, 1)
        x = self.fc1(x)
        x = F.relu(x)
        x = self.dropout2(x)
        x = self.fc2(x)
        output = F.log_softmax(x, dim=1)
        return output

# Training Function
def train(args, model, device, train_loader, optimizer, epoch):
    model.train()
    for batch_idx, (data, target) in enumerate(train_loader):
        data, target = data.to(device), target.to(device)
        optimizer.zero_grad()
        output = model(data)
        loss = F.nll_loss(output, target)
        loss.backward()
        optimizer.step()
        if batch_idx % args.log_interval == 0:
            print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
                epoch, batch_idx * len(data), len(train_loader.dataset),
                100. * batch_idx / len(train_loader), loss.item()))
            if args.dry_run:
                break

# Testing Function
def test(model, device, test_loader):
    model.eval()
    test_loss = 0
    correct = 0
    with torch.no_grad():
        for data, target in test_loader:
            data, target = data.to(device), target.to(device)
            output = model(data)
            test_loss += F.nll_loss(output, target, reduction='sum').item()
            pred = output.argmax(dim=1, keepdim=True)
            correct += pred.eq(target.view_as(pred)).sum().item()

    test_loss /= len(test_loader.dataset)

    print('\nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.0f}%)\n'.format(
        test_loss, correct, len(test_loader.dataset),
        100. * correct / len(test_loader.dataset)))

# SHAP Explanation Function
def explain_with_shap(model, device, test_loader):
    # Put model in evaluation mode
    model.eval()

    # Select a small subset of test data for SHAP explanations
    data_iter = iter(test_loader)
    data, _ = next(data_iter)  # Get a single batch of test data
    data = data[:10].to(device)  # Limit to 10 samples for visualization

    print("Generating SHAP explanations...")

    # Create SHAP Explainer
    explainer = shap.DeepExplainer(model, data)  # Use the test batch as the background
    shap_values = explainer.shap_values(data)  # Calculate SHAP values for the batch

    # Visualize explanations
    for i in range(len(data)):
        img = data[i].cpu().numpy().squeeze()  # Convert tensor to numpy array
        shap_img = np.sum(shap_values[0][i], axis=0)  # Aggregate SHAP values over output channels

        # Plot the original image and SHAP explanation
        plt.figure(figsize=(10, 5))
        plt.subplot(1, 2, 1)
        plt.title("Original Image")
        plt.imshow(img, cmap='gray')
        plt.subplot(1, 2, 2)
        plt.title("SHAP Explanation")
        plt.imshow(shap_img, cmap='coolwarm')
        plt.colorbar(label="Pixel Importance")
        plt.show()

def main():
    # Parse command-line arguments
    parser = argparse.ArgumentParser(description='PyTorch MNIST Example with SHAP')
    parser.add_argument('--batch-size', type=int, default=64, metavar='N',
                        help='input batch size for training (default: 64)')
    parser.add_argument('--test-batch-size', type=int, default=1000, metavar='N',
                        help='input batch size for testing (default: 1000)')
    parser.add_argument('--epochs', type=int, default=10, metavar='N',
                        help='number of epochs to train (default: 14)')
    parser.add_argument('--lr', type=float, default=1.0, metavar='LR',
                        help='learning rate (default: 1.0)')
    parser.add_argument('--gamma', type=float, default=0.7, metavar='M',
                        help='Learning rate step gamma (default: 0.7)')
    parser.add_argument('--no-cuda', action='store_true', default=False,
                        help='disables CUDA training')
    parser.add_argument('--dry-run', action='store_true', default=False,
                        help='quickly check a single pass')
    parser.add_argument('--seed', type=int, default=1, metavar='S',
                        help='random seed (default: 1)')
    parser.add_argument('--log-interval', type=int, default=10, metavar='N',
                        help='how many batches to wait before logging training status')
    parser.add_argument('--save-model', action='store_true', default=False,
                        help='For Saving the current Model')
    args = parser.parse_args()
    use_cuda = not args.no_cuda and torch.cuda.is_available()

    torch.manual_seed(args.seed)

    # Device configuration
    device = torch.device("mps" if torch.mps.is_available() else "cpu")
    if torch.mps.is_available():
        print('Using MPS')

    # Data loading and preprocessing
    train_kwargs = {'batch_size': args.batch_size}
    test_kwargs = {'batch_size': args.test_batch_size}
    if use_cuda:
        cuda_kwargs = {'num_workers': 1,
                       'pin_memory': True,
                       'shuffle': True}
        train_kwargs.update(cuda_kwargs)
        test_kwargs.update(cuda_kwargs)

    transform = transforms.Compose([
        transforms.ToTensor(),
        transforms.Normalize((0.1307,), (0.3081,))
    ])
    dataset1 = datasets.MNIST('../data', train=True, download=True,
                               transform=transform)
    dataset2 = datasets.MNIST('../data', train=False,
                               transform=transform)
    train_loader = torch.utils.data.DataLoader(dataset1, **train_kwargs)
    test_loader = torch.utils.data.DataLoader(dataset2, **test_kwargs)

    # Initialize model, optimizer, and learning rate scheduler
    model = Net().to(device)
    optimizer = optim.Adadelta(model.parameters(), lr=args.lr)
    scheduler = StepLR(optimizer, step_size=1, gamma=args.gamma)

    # Training and testing loop
    for epoch in range(1, args.epochs + 1):
        train(args, model, device, train_loader, optimizer, epoch)
        test(model, device, test_loader)
        scheduler.step()

    # Save the trained model
    if args.save_model:
        torch.save(model.state_dict(), "mnist_cnn.pt")

    # SHAP explanations
    explain_with_shap(model, device, test_loader)


if __name__ == '__main__':
    main()