Mlforecast

Bug Fixes

- fix(compat): handle zero offset in shift_array jmoralez (480)
- fix(global-sklearn-tfm): apply inverse transform to each column jmoralez (477)

Bug Fixes

- fix: X_df handling in direct approach jmoralez (468)
- fix(auto): remove invalid params from xgboost default space jmoralez (464)

Breaking Change

- breaking: remove window_ops and numba dependencies jmoralez (462)

Changes
- chore: deprecate window_ops jmoralez (410)

New Features

- feat(distributed): support ids in predict jmoralez (454)
- feat(auto): support input_size jmoralez (451)

Breaking Change

- breaking: drop rows with null targets when `dropna=False` jmoralez (447)

Bug Fixes

- fix(auto): support custom column names jmoralez (449)

Enhancement

- enh(distributed): propagate null features in spark jmoralez (448)

New Features

- feat: add `weight_col` to `MLForecast.fit` and `MLForecast.cross_validation` jmoralez (444)
- feat: infer samples required for built-in lag transforms updates jmoralez (445)