def update()

in level_replay/algo/ppo.py [0:0]
48 lines of code
4 McCabe index (conditional complexity)

    def update(self, rollouts):
        advantages = rollouts.returns[:-1] - rollouts.value_preds[:-1]
        advantages = (advantages - advantages.mean()) / (
            advantages.std() + 1e-5)

        value_loss_epoch = 0
        action_loss_epoch = 0
        dist_entropy_epoch = 0

        for e in range(self.ppo_epoch):
            if self.actor_critic.is_recurrent:
                data_generator = rollouts.recurrent_generator(
                    advantages, self.num_mini_batch)
            else:
                data_generator = rollouts.feed_forward_generator(
                    advantages, self.num_mini_batch)

            for sample in data_generator:
                obs_batch, recurrent_hidden_states_batch, actions_batch, \
                value_preds_batch, return_batch, masks_batch, old_action_log_probs_batch, \
                        adv_targ = sample
                
                values, action_log_probs, dist_entropy, _ = self.actor_critic.evaluate_actions(
                    obs_batch, recurrent_hidden_states_batch, masks_batch,
                    actions_batch)
                    
                ratio = torch.exp(action_log_probs -
                                  old_action_log_probs_batch)
                surr1 = ratio * adv_targ
                surr2 = torch.clamp(ratio, 1.0 - self.clip_param,
                                    1.0 + self.clip_param) * adv_targ
                action_loss = -torch.min(surr1, surr2).mean()

                value_pred_clipped = value_preds_batch + \
                    (values - value_preds_batch).clamp(-self.clip_param, self.clip_param)
                value_losses = (values - return_batch).pow(2)
                value_losses_clipped = (
                    value_pred_clipped - return_batch).pow(2)
                value_loss = 0.5 * torch.max(value_losses,
                                                value_losses_clipped).mean()

                self.optimizer.zero_grad()
                loss = (value_loss*self.value_loss_coef + action_loss - dist_entropy*self.entropy_coef)
                loss.backward()
                nn.utils.clip_grad_norm_(self.actor_critic.parameters(),
                                        self.max_grad_norm)
                self.optimizer.step()  
                                
                value_loss_epoch += value_loss.item()
                action_loss_epoch += action_loss.item()
                dist_entropy_epoch += dist_entropy.item()

        num_updates = self.ppo_epoch * self.num_mini_batch

        value_loss_epoch /= num_updates
        action_loss_epoch /= num_updates
        dist_entropy_epoch /= num_updates

        return value_loss_epoch, action_loss_epoch, dist_entropy_epoch