修复反归一化错误

config/REPST/BeijingAirQuality.yaml

@@ -11,8 +11,8 @@ data:
   column_wise: false
   days_per_week: 7
   default_graph: true
-  horizon: 12
-  lag: 12
+  horizon: 24
+  lag: 24
   normalizer: std
   num_nodes: 7
   steps_per_day: 288
@@ -24,8 +24,8 @@ data:
   batch_size: 16
 
 model:
-  pred_len: 12
-  seq_len: 12
+  pred_len: 24
+  seq_len: 24
   patch_len: 6 
   stride: 7
   dropout: 0.2

config/REPST/METR-LA.yaml

@@ -11,8 +11,8 @@ data:
   column_wise: false
   days_per_week: 7
   default_graph: true
-  horizon: 12
-  lag: 12
+  horizon: 24
+  lag: 24
   normalizer: std
   num_nodes: 207
   steps_per_day: 288
@@ -24,8 +24,8 @@ data:
   batch_size: 16
 
 model:
-  pred_len: 12
-  seq_len: 12
+  pred_len: 24
+  seq_len: 24
   patch_len: 6 
   stride: 7
   dropout: 0.2
@@ -41,7 +41,7 @@ train:
   batch_size: 16
   early_stop: true
   early_stop_patience: 15
-  epochs: 100
+  epochs: 1
   grad_norm: false
   loss_func: mae
   lr_decay: true
@@ -52,7 +52,7 @@ train:
   real_value: true
   weight_decay: 0
   debug: false
-  output_dim: 1
+  output_dim: 100
   log_step: 1000
   plot: false
   mae_thresh: None

config/REPST/SolarEnergy.yaml

@@ -11,8 +11,8 @@ data:
   column_wise: false
   days_per_week: 7
   default_graph: true
-  horizon: 12
-  lag: 12
+  horizon: 24
+  lag: 24
   normalizer: std
   num_nodes: 137
   steps_per_day: 288
@@ -24,8 +24,8 @@ data:
   batch_size: 16
 
 model:
-  pred_len: 12
-  seq_len: 12
+  pred_len: 24
+  seq_len: 24
   patch_len: 6 
   stride: 7
   dropout: 0.2

model/REPST/normalizer.py

@@ -13,9 +13,7 @@ class GumbelSoftmax(nn.Module):
         return self.gumbel_softmax(logits, 1, self.k, self.hard)
 
     def gumbel_softmax(self, logits, tau=1, k=1000, hard=True):
-
         y_soft = F.gumbel_softmax(logits, tau, hard)
-
         if hard:
             # 生成硬掩码
             _, indices = y_soft.topk(k, dim=0)  # 选择Top-K

model/REPST/reprogramming.py

@@ -51,7 +51,6 @@ class PatchEmbedding(nn.Module):
         x = self.padding_patch_layer(x)
         x = x.unfold(dimension=-1, size=self.patch_len, step=self.stride)
         x_value_embed = self.value_embedding(x)
-
         return self.dropout(x_value_embed), n_vars
     
 class ReprogrammingLayer(nn.Module):
@@ -83,13 +82,9 @@ class ReprogrammingLayer(nn.Module):
 
     def reprogramming(self, target_embedding, source_embedding, value_embedding):
         B, L, H, E = target_embedding.shape
-
         scale = 1. / sqrt(E)
-
         scores = torch.einsum("blhe,she->bhls", target_embedding, source_embedding)
-
         A = self.dropout(torch.softmax(scale * scores, dim=-1))
         reprogramming_embedding = torch.einsum("bhls,she->blhe", A, value_embedding)
-
         return reprogramming_embedding
     

trainer/Trainer.py

@@ -204,8 +204,8 @@ class Trainer:
 
                 # 累积损失和预测结果
                 total_loss += loss.item()
-                y_pred.append(output.detach().cpu())
-                y_true.append(label.detach().cpu())
+                y_pred.append(d_output.detach().cpu())
+                y_true.append(d_label.detach().cpu())
 
                 # 更新进度条
                 progress_bar.set_postfix(loss=d_loss.item())
@@ -356,18 +356,15 @@ class Trainer:
                 y_pred.append(output)
                 y_true.append(label)
 
-        # 合并所有批次的预测结果
-        if args["real_value"]:
-            y_pred = scaler.inverse_transform(torch.cat(y_pred, dim=0))
-        else:
-            y_pred = torch.cat(y_pred, dim=0)
-        y_true = torch.cat(y_true, dim=0)
+        
+        d_y_pred = scaler.inverse_transform(torch.cat(y_pred, dim=0))
+        d_y_true = scaler.inverse_transform(torch.cat(y_true, dim=0))
 
         # 计算并记录每个时间步的指标
-        for t in range(y_true.shape[1]):
+        for t in range(d_y_true.shape[1]):
             mae, rmse, mape = all_metrics(
-                y_pred[:, t, ...],
-                y_true[:, t, ...],
+                d_y_pred[:, t, ...],
+                d_y_true[:, t, ...],
                 args["mae_thresh"],
                 args["mape_thresh"],
             )
@@ -377,7 +374,7 @@ class Trainer:
 
         # 计算并记录平均指标
         mae, rmse, mape = all_metrics(
-            y_pred, y_true, args["mae_thresh"], args["mape_thresh"]
+            d_y_pred, d_y_true, args["mae_thresh"], args["mape_thresh"]
         )
         logger.info(
             f"Average Horizon, MAE: {mae:.4f}, RMSE: {rmse:.4f}, MAPE: {mape:.4f}"