import torch
import torch.nn as nn

class KL_Loss(nn.Module):
    def __init__(self, logvar_init=0.0, kl_weight=1.0, nll_loss_type='l1'):

        super().__init__()
        self.kl_weight = kl_weight
        # output log variance
        self.logvar = nn.Parameter(torch.ones(size=()) * logvar_init)
        if nll_loss_type == 'l1':
            self.nll_loss = torch.nn.L1Loss(reduction='none')
        elif nll_loss_type == 'l2':
            self.nll_loss = torch.nn.MSELoss(reduction='none')
        else:
            self.nll_loss = None
            # give a warning
            print("Warning: nll_loss_type not recognized, no nll_loss will be used.")
    def calculate_adaptive_weight(self, nll_loss, g_loss, last_layer=None):
        if last_layer is not None:
            nll_grads = torch.autograd.grad(nll_loss, last_layer, retain_graph=True)[0]
            g_grads = torch.autograd.grad(g_loss, last_layer, retain_graph=True)[0]
        else:
            nll_grads = torch.autograd.grad(nll_loss, self.last_layer[0], retain_graph=True)[0]
            g_grads = torch.autograd.grad(g_loss, self.last_layer[0], retain_graph=True)[0]

        d_weight = torch.norm(nll_grads) / (torch.norm(g_grads) + 1e-4)
        d_weight = torch.clamp(d_weight, 0.0, 1e4).detach()
        d_weight = d_weight * self.discriminator_weight
        return d_weight

    def forward(self, inputs, reconstructions, posteriors, split="train", weights=None):
        if self.nll_loss is None:
            rec_loss = torch.tensor(0.0)
            nll_loss = torch.tensor(0.0)
        else:
            rec_loss = self.nll_loss(inputs.contiguous(), reconstructions.contiguous())

            nll_loss = rec_loss / torch.exp(self.logvar) + self.logvar
            weighted_nll_loss = nll_loss
            if weights is not None:
                weighted_nll_loss = weights*nll_loss
            weighted_nll_loss = torch.sum(weighted_nll_loss) / weighted_nll_loss.shape[0]
            nll_loss = torch.sum(nll_loss) / nll_loss.shape[0]
        
        kl_loss = posteriors.kl()
        kl_loss = torch.sum(kl_loss) / kl_loss.shape[0]
        
        if self.nll_loss is None:
            loss = self.kl_weight * kl_loss
        else:
            loss = weighted_nll_loss + self.kl_weight * kl_loss

        log = {"{}/total_loss".format(split): loss.clone().detach().mean(), "{}/logvar".format(split): self.logvar.detach(),
                "{}/kl_loss".format(split): kl_loss.detach().mean(), "{}/nll_loss".format(split): nll_loss.detach().mean(),
                "{}/rec_loss".format(split): rec_loss.detach().mean(),
                }
        return loss, log