Fix item_id for M5 dataset (#3156)

*Description of changes:*
- Fix the `item_id` calculation for the M5 dataset in case `id` column
is missing in the original dataset

By submitting this pull request, I confirm that you can use, modify,
copy, and redistribute this contribution, under the terms of your
choice.


**Please tag this pr with at least one of these labels to make our
release process faster:** BREAKING, new feature, bug fix, other change,
dev setup

src/gluonts/dataset/repository/_m5.py

@@ -112,16 +112,18 @@ def generate_m5_dataset(
  len(state_ids_un),
  ]
 
+ # Compute unique ID in case `id` column is missing
+ if "id" not in sales_train_validation.columns:
+ sales_train_validation["id"] = (
+ sales_train_validation["item_id"].astype("str")
+ + "_"
+ + sales_train_validation["store_id"].astype("str")
+ )
  # Build target series
- train_ids = (
- sales_train_validation["item_id"].str
- + "_"
- + sales_train_validation["store_id"].str
- )
+ train_ids = sales_train_validation["id"]
  train_df = sales_train_validation.drop(
  ["id", "item_id", "dept_id", "cat_id", "store_id", "state_id"],
  axis=1,
- errors="ignore",
  )
  test_target_values = train_df.values.copy()
  train_target_values = [ts[:-prediction_length] for ts in train_df.values]