fix item_id field in provided datasets (#566)

src/gluonts/dataset/repository/_gp_copula_2019.py

@@ -147,6 +147,7 @@ def save_dataset(dataset_path: Path, ds_info: GPCopulaDataset):
  # Handles adding categorical features of rolling
  # evaluation dates
  cat=[cat - ds_info.num_series * (cat // ds_info.num_series)],
+ item_id=cat,
  )
  for cat, data_entry in enumerate(dataset)
  ],

src/gluonts/dataset/repository/_lstnet.py

@@ -166,6 +166,7 @@ def generate_lstnet_dataset(dataset_path: Path, dataset_name: str):
  target_values=sliced_ts.values,
  start=sliced_ts.index[0],
  cat=[cat],
+ item_id=cat,
  )
  )
 
@@ -192,6 +193,7 @@ def generate_lstnet_dataset(dataset_path: Path, dataset_name: str):
  target_values=sliced_ts.values,
  start=sliced_ts.index[0],
  cat=[cat],
+ item_id=cat,
  )
  )
 

src/gluonts/dataset/repository/_m4.py

@@ -72,15 +72,25 @@ def generate_m4_dataset(
  save_to_file(
  train_file,
  [
- to_dict(target_values=target, start=mock_start_dataset, cat=[cat])
+ to_dict(
+ target_values=target,
+ start=mock_start_dataset,
+ cat=[cat],
+ item_id=cat,
+ )
  for cat, target in enumerate(train_target_values)
  ],
  )
 
  save_to_file(
  test_file,
  [
- to_dict(target_values=target, start=mock_start_dataset, cat=[cat])
+ to_dict(
+ target_values=target,
+ start=mock_start_dataset,
+ cat=[cat],
+ item_id=cat,
+ )
  for cat, target in enumerate(test_target_values)
  ],
  )

src/gluonts/dataset/repository/_util.py

@@ -14,13 +14,16 @@
 import json
 import os
 from pathlib import Path
-from typing import Dict, List, Optional
+from typing import Dict, List, Optional, Any
 
 import numpy as np
 
 
 def to_dict(
- target_values: np.ndarray, start: str, cat: Optional[List[int]] = None
+ target_values: np.ndarray,
+ start: str,
+ cat: Optional[List[int]] = None,
+ item_id: Optional[Any] = None,
 ):
  def serialize(x):
  if np.isnan(x):
@@ -37,6 +40,9 @@ def serialize(x):
  if cat is not None:
  res["feat_static_cat"] = cat
 
+ if item_id is not None:
+ res["item_id"] = item_id
+
  return res