import torch
import torch.nn as nn
import torch.optim as optim
from torchvision import datasets, transforms
from torch.utils.data import DataLoader

class MLPRationalTerm(nn.Module):
    def __init__(self, input_dim, hidden_dim, output_dim):
        super(MLPRationalTerm, self).__init__()
        self.net = nn.Sequential(
            nn.Linear(input_dim, hidden_dim),
            nn.ReLU(),
            nn.Linear(hidden_dim, output_dim)
        )

    def forward(self, x):
        return self.net(x)

class nRANClassifier(nn.Module):
    def __init__(self, n_terms, input_dim=784, hidden_dim=32, output_dim=10):
        super(nRANClassifier, self).__init__()
        self.n_terms = n_terms
        self.base_offset = nn.Linear(input_dim, output_dim)
        self.rational_terms = nn.ModuleList([
            MLPRationalTerm(input_dim, hidden_dim, output_dim) for _ in range(n_terms)
        ])
        self.scales = [0.1 ** i for i in range(1, n_terms + 1)]

    def forward(self, x):
        x = x.view(x.size(0), -1)
        out = self.base_offset(x)
        for i, term in enumerate(self.rational_terms):
            out += self.scales[i] * term(x)
        return out

def train_test_mnist():
    device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
    print(f"Using device: {device}")
    transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.1307,), (0.3081,))])
    train_loader = DataLoader(datasets.MNIST('./data', train=True, download=True, transform=transform), batch_size=64, shuffle=True)
    test_loader = DataLoader(datasets.MNIST('./data', train=False, transform=transform), batch_size=1000)

    model = nRANClassifier(n_terms=5).to(device)
    optimizer = optim.Adam(model.parameters(), lr=0.001)
    criterion = nn.CrossEntropyLoss()

    model.train()
    for epoch in range(1, 3):
        for batch_idx, (data, target) in enumerate(train_loader):
            data, target = data.to(device), target.to(device)
            optimizer.zero_grad()
            output = model(data)
            loss = criterion(output, target)
            loss.backward()
            optimizer.step()
            if batch_idx % 200 == 0:
                print(f"Epoch {epoch} [{batch_idx*64}/60000] Loss: {loss.item():.6f}")

    model.eval()
    correct = 0
    with torch.no_grad():
        for data, target in test_loader:
            data, target = data.to(device), target.to(device)
            output = model(data)
            pred = output.argmax(dim=1, keepdim=True)
            correct += pred.eq(target.view_as(pred)).sum().item()

    print(f"\nTest set Accuracy: {correct}/{len(test_loader.dataset)} ({100. * correct / len(test_loader.dataset):.2f}%)")

if __name__ == '__main__':
    train_test_mnist()