更新iTransformer, HI配置。更新TS数据集载入方式

config/HI/AirQuality.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 16
+  batch_size: 512
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -25,7 +25,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 512
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -40,7 +40,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 35
+  output_dim: 6
   optimizer: null
   plot: false
   real_value: true

config/HI/BJTaxi-Inflow.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 32
+  batch_size: 2048
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -25,7 +25,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 2048
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -40,7 +40,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 1024
+  output_dim: 1
   optimizer: null
   plot: false
   real_value: true

config/HI/BJTaxi-Outflow.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 32
+  batch_size: 2048
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -25,7 +25,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 2048
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -40,7 +40,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 1024
+  output_dim: 1
   optimizer: null
   plot: false
   real_value: true

config/HI/METR-LA.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 16
+  batch_size: 512
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -25,7 +25,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 512
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -40,7 +40,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 207
+  output_dim: 1
   optimizer: null
   plot: false
   real_value: true

config/HI/NYCBike-Inflow.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 32
+  batch_size: 512
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -25,7 +25,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 512
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -40,7 +40,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 128
+  output_dim: 1
   optimizer: null
   plot: false
   real_value: true

config/HI/NYCBike-Outflow.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 32
+  batch_size: 512
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -25,7 +25,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 512
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -40,7 +40,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 128
+  output_dim: 1
   optimizer: null
   plot: false
   real_value: true

config/HI/PEMS-BAY.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 16
+  batch_size: 512
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -25,7 +25,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 512
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -40,7 +40,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 325
+  output_dim: 1
   optimizer: null
   plot: false
   real_value: true

config/HI/SolarEnergy.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 16
+  batch_size: 512
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -25,7 +25,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 512
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -40,7 +40,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 137
+  output_dim: 1
   optimizer: null
   plot: false
   real_value: true

config/iTransformer/AirQuality.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 16
+  batch_size: 256
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -31,7 +31,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 256
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -46,7 +46,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 35
+  output_dim: 6
   plot: false
   real_value: true
   weight_decay: 0

config/iTransformer/BJTaxi-Inflow.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 32
+  batch_size: 2048
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -31,7 +31,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 2048
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -46,7 +46,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 1024
+  output_dim: 1
   plot: false
   real_value: true
   weight_decay: 0

config/iTransformer/BJTaxi-Outflow.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 32
+  batch_size: 2048
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -31,7 +31,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 2048
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -46,7 +46,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 1024
+  output_dim: 1
   plot: false
   real_value: true
   weight_decay: 0

config/iTransformer/METR-LA.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 16
+  batch_size: 256
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -31,7 +31,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 256
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -46,7 +46,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 207
+  output_dim: 1
   plot: false
   real_value: true
   weight_decay: 0

config/iTransformer/NYCBike-Inflow.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 32
+  batch_size: 256
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -31,7 +31,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 256
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -46,7 +46,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 128
+  output_dim: 1
   plot: false
   real_value: true
   weight_decay: 0

config/iTransformer/NYCBike-Outflow.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 32
+  batch_size: 256
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -31,7 +31,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 256
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -46,7 +46,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 128
+  output_dim: 1
   plot: false
   real_value: true
   weight_decay: 0

config/iTransformer/PEMS-BAY.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 16
+  batch_size: 256
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -31,7 +31,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 256
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -46,7 +46,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 325
+  output_dim: 1
   plot: false
   real_value: true
   weight_decay: 0

config/iTransformer/SolarEnergy.yaml

@@ -6,7 +6,7 @@ basic:
   seed: 2023
 
 data:
-  batch_size: 16
+  batch_size: 256
   column_wise: false
   days_per_week: 7
   horizon: 24
@@ -31,7 +31,7 @@ model:
 
 
 train:
-  batch_size: 16
+  batch_size: 256
   debug: false
   early_stop: true
   early_stop_patience: 15
@@ -46,7 +46,7 @@ train:
   mae_thresh: None
   mape_thresh: 0.001
   max_grad_norm: 5
-  output_dim: 137
+  output_dim: 1
   plot: false
   real_value: true
   weight_decay: 0

dataloader/TSloader.py

@@ -7,16 +7,16 @@ import torch
 
 def get_dataloader(args, normalizer="std", single=True):
     data = load_st_dataset(args)
-    data = data[..., 0:1]
+    # data = data[..., 0:1]
 
     args = args["data"]
     L, N, F = data.shape
-    data = data.reshape(L, N*F)  # [L, N*F]
+    # data = data.reshape(L, N*F)  # [L, N*F]
 
     # Generate sliding windows for main data and add time features
     x, y = _prepare_data_with_windows(data, args, single)
     
-    # Split data
+    # Split data [b,t,n,c]
     split_fn = split_data_by_days if args["test_ratio"] > 1 else split_data_by_ratio
     x_train, x_val, x_test = split_fn(x, args["val_ratio"], args["test_ratio"])
     y_train, y_val, y_test = split_fn(y, args["val_ratio"], args["test_ratio"])
@@ -25,6 +25,10 @@ def get_dataloader(args, normalizer="std", single=True):
     scaler = _normalize_data(x_train, x_val, x_test, args, normalizer)
     _apply_existing_scaler(y_train, y_val, y_test, scaler, args)
 
+    # reshape [b,t,n,c] -> [b*n, t, c]
+    x_train, x_val, x_test, y_train, y_val, y_test = \
+        _reshape_tensor(x_train, x_val, x_test, y_train, y_val, y_test)
+
     # Create dataloaders
     return (
         _create_dataloader(x_train, y_train, args["batch_size"], True, False),
@@ -33,6 +37,16 @@ def get_dataloader(args, normalizer="std", single=True):
         scaler
     )
 
+def _reshape_tensor(*tensors):
+    """Reshape tensors from [b, t, n, c] -> [b*n, t, c]."""
+    reshaped = []
+    for x in tensors:
+        # x 是 ndarray：shape (b, t, n, c)
+        b, t, n, c = x.shape
+        x_new = x.transpose(0, 2, 1, 3).reshape(b * n, t, c)
+        reshaped.append(x_new)
+    return reshaped
+
 def _prepare_data_with_windows(data, args, single):
     # Generate sliding windows for main data
     x = add_window_x(data, args["lag"], args["horizon"], single)