Auto-Create Timestamps in prettify_prediction() When test_data is None

[aditisingh02]

Why are these changes needed?

Currently, the TimeSeriesDataset.prettify_prediction() method in flaml/automl/time_series/ts_data.py throws a NotImplementedError when test_data is None:

# TODO auto-create the timestamps for the time column instead of throwing
raise NotImplementedError("Need a non-None test_data for this to work, for now")

This is frustrating for users who want to make predictions without providing explicit test data timestamps, which is a common use case in time series forecasting.

This PR implements automatic timestamp generation by:

1. Using the training data's end date (train_data[time_col].max()) as the starting point
2. Generating future timestamps based on the dataset's inferred frequency
3. Supporting all input types: np.ndarray, pd.Series, and pd.DataFrame

Example behavior after this change:

# Before: NotImplementedError
# After: Automatically generates timestamps starting from training end + 1 period
y_pred = model.predict(steps=10)  # Works without explicit test_data!

Related issue number

Closes #1506

Checks

• I've used pre-commit to lint the changes in this PR (black and ruff passed; pyupgrade failed due to Python 3.14 ast module incompatibility - not a code issue).
• I've included any doc changes needed for https://microsoft.github.io/FLAML/. (No doc changes needed - internal implementation fix)
• I've added tests (if relevant) corresponding to the changes introduced in this PR.
• I've made sure all auto checks have passed.