main.py

import matplotlib.pyplot as plt
import numpy as np

from config import load_args_contrastive
from dataset_ref import load_data_duo
import torch
import torch.nn as nn
from model import Classification_model_duo_contrastive
import torch.optim as optim
from sklearn.metrics import confusion_matrix
import seaborn as sn
import pandas as pd

def train_duo(model, data_train, optimizer, loss_function, epoch):
    model.train()
    losses = 0.
    acc = 0.
    for param in model.parameters():
        param.requires_grad = True

    for imaer,imana, img_ref, label in data_train:
        label = label.long()
        if torch.cuda.is_available():
            imaer = imaer.cuda()
            imana = imana.cuda()
            img_ref = img_ref.cuda()
            label = label.cuda()
        pred_logits = model.forward(imaer,imana,img_ref)
        pred_class = torch.argmax(pred_logits,dim=1)
        acc += (pred_class==label).sum().item()
        loss = loss_function(pred_logits,label)
        losses += loss.item()
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
    losses = losses/len(data_train.dataset)
    acc = acc/len(data_train.dataset)
    print('Train epoch {}, loss : {:.3f} acc : {:.3f}'.format(epoch,losses,acc))
    return losses, acc

def test_duo(model, data_test, loss_function, epoch):
    model.eval()
    losses = 0.
    acc = 0.
    acc_contrastive = 0.
    for param in model.parameters():
        param.requires_grad = False

    for imaer,imana, img_ref, label in data_test:
        imaer = imaer.transpose(0,1)
        imana = imana.transpose(0,1)
        img_ref = img_ref.transpose(0,1)
        label = label.transpose(0,1)
        label = label.squeeze()
        label = label.long()
        if torch.cuda.is_available():
            imaer = imaer.cuda()
            imana = imana.cuda()
            img_ref = img_ref.cuda()
            label = label.cuda()
        label_class = torch.argmin(label).data.cpu().numpy()
        pred_logits = model.forward(imaer,imana,img_ref)
        pred_class = torch.argmax(pred_logits[:,0]).tolist()
        acc_contrastive += (torch.argmax(pred_logits,dim=1).data.cpu().numpy()==label.data.cpu().numpy()).sum().item()
        acc += (pred_class==label_class)
        loss = loss_function(pred_logits,label)
        losses += loss.item()
    losses = losses/(label.shape[0]*len(data_test.dataset))
    acc = acc/(len(data_test.dataset))
    acc_contrastive = acc_contrastive /(label.shape[0]*len(data_test.dataset))
    print('Test epoch {}, loss : {:.3f}  acc : {:.3f} acc contrastive : {:.3f}'.format(epoch,losses,acc,acc_contrastive))
    return losses,acc,acc_contrastive

def run_duo(args):
    #load data
    data_train, data_test_batch = load_data_duo(base_dir_train=args.dataset_train_dir, base_dir_test=args.dataset_val_dir, batch_size=args.batch_size,
                                          ref_dir=args.dataset_ref_dir, positive_prop=args.positive_prop)

    #load model
    model = Classification_model_duo_contrastive(model = args.model, n_class=2)
    model.double()
    #load weight
    if args.pretrain_path is not None :
        load_model(model,args.pretrain_path)
    #move parameters to GPU
    if torch.cuda.is_available():
        model = model.cuda()

    #init accumulators
    best_loss = 100
    train_acc=[]
    train_loss=[]
    val_acc=[]
    val_cont_acc=[]
    val_loss=[]
    #init training
    loss_function = nn.CrossEntropyLoss()
    optimizer = optim.SGD(model.parameters(), lr=0.001, momentum=0.9)
    #train model
    for e in range(args.epoches):
        loss, acc = train_duo(model,data_train,optimizer,loss_function,e)
        train_loss.append(loss)
        train_acc.append(acc)
        if e%args.eval_inter==0 :
            loss, acc, acc_contrastive = test_duo(model,data_test_batch,loss_function,e)
            val_loss.append(loss)
            val_acc.append(acc)
            val_cont_acc.append(acc_contrastive)
            if loss < best_loss :
                save_model(model,args.save_path)
                best_loss = loss
    # plot and save training figs
    plt.clf()
    plt.subplot(2, 1, 1)
    plt.plot(train_acc, label='train cont acc')
    plt.plot(val_cont_acc, label='val cont acc')
    plt.plot(val_acc, label='val classification acc')
    plt.title('Train and validation accuracy')
    plt.xlabel('epoch')
    plt.ylabel('accuracy')
    plt.legend(loc="upper left")
    plt.ylim(0, 1.05)
    plt.tight_layout()

    plt.subplot(2, 1, 2)
    plt.plot(train_loss, label='train')
    plt.plot(val_loss, label='val')
    plt.title('Train and validation loss')
    plt.xlabel('epoch')
    plt.ylabel('loss')
    plt.legend(loc="upper left")
    plt.tight_layout()


    plt.show()
    plt.savefig('output/training_plot_contrastive_noise_{}_lr_{}_model_{}.png'.format(args.noise_threshold,args.lr,args.model))

    #load and evaluate best model
    load_model(model, args.save_path)
    make_prediction_duo(model,data_test_batch, 'output/confusion_matrix_contractive_noise_{}_lr_{}_model_{}.png'.format(args.noise_threshold,args.lr,args.model),
                        'output/confidence_matrix_contractive_noise_{}_lr_{}_model_{}.png'.format(args.noise_threshold,args.lr,args.model))


def make_prediction_duo(model, data, f_name, f_name2):
    for imaer, imana, img_ref, label in data:
        n_class = label.shape[1]
        break
    confidence_pred_list = [[] for i in range(n_class)]
    y_pred = []
    y_true = []
    soft_max = nn.Softmax(dim=1)
    # iterate over test data
    for imaer,imana,img_ref, label in data:
        imaer = imaer.transpose(0,1)
        imana = imana.transpose(0,1)
        img_ref = img_ref.transpose(0,1)
        label = label.transpose(0,1)
        label = label.squeeze()
        label = label.long()
        specie = torch.argmin(label)


        if torch.cuda.is_available():
            imaer = imaer.cuda()
            imana = imana.cuda()
            img_ref = img_ref.cuda()
            label = label.cuda()
        output = model(imaer,imana,img_ref)
        confidence = soft_max(output)
        confidence_pred_list[specie].append(confidence[:,0].data.cpu().numpy())
        #Mono class output (only most postive paire)
        output = torch.argmax(output[:,0])
        label = torch.argmin(label)
        y_pred.append(output.tolist())
        y_true.append(label.tolist())  # Save Truth
    # constant for classes

    # Build confusion matrix
    classes = data.dataset.classes
    cf_matrix = confusion_matrix(y_true, y_pred)
    confidence_matrix = np.zeros((n_class,n_class))
    for i in range(n_class):
        confidence_matrix[i]=np.mean(confidence_pred_list[i],axis=0)

    df_cm = pd.DataFrame(cf_matrix / np.sum(cf_matrix, axis=1)[:, None], index=[i for i in classes],
                         columns=[i for i in classes])
    print('Saving Confusion Matrix')
    plt.clf()
    plt.figure(figsize=(14, 9))
    sn.heatmap(df_cm, annot=cf_matrix)
    plt.savefig(f_name)

    df_cm = pd.DataFrame(confidence_matrix, index=[i for i in classes],
                         columns=[i for i in classes])
    print('Saving Confidence Matrix')
    plt.clf()
    plt.figure(figsize=(14, 9))
    sn.heatmap(df_cm, annot=confidence_matrix)
    plt.savefig(f_name2)


def save_model(model, path):
    print('Model saved')
    torch.save(model.state_dict(), path)

def load_model(model, path):
    model.load_state_dict(torch.load(path, weights_only=True))



if __name__ == '__main__':
    args = load_args_contrastive()
    run_duo(args)