[ML-52572] Refactor and add tests for covariate inference support in automl runtime for deepar

[Pajaraja]

Add comments to code explaining covariate support in deepAR.

Add new tests for changes to function set_index_and_fill_missing_time_steps for covariates.

[Copilot]

Pull Request Overview

This PR adds documentation and test coverage for covariate support in the DeepAR forecasting model. The changes clarify how covariates are handled during preprocessing, particularly around forward/backward filling and alignment with timestamps.

Key changes:

• Added explanatory comments in utils.py and model.py describing covariate handling logic
• Added two new test cases validating covariate preprocessing for both single and multi-series scenarios

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
runtime/tests/automl_runtime/forecast/deepar/utils_test.py	Added two new test methods to verify covariate forward/backward filling and alignment for uni- and multi-timeseries cases
runtime/databricks/automl_runtime/forecast/deepar/utils.py	Enhanced comments explaining covariate filling behavior and rationale for not filling target column
runtime/databricks/automl_runtime/forecast/deepar/model.py	Added comments documenting preprocessing steps, covariate aggregation strategy, and GluonTS dataset format requirements

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}