Changes in version 1.1.0 (2025-09-04)                  

  - Major update to align with tune 2.0.0.
    
      - Updated internal logic for compatibility with the new column
        naming conventions in resampling results (.model_desc vs .row).
      - Improved handling of .resample_id and .row_id to ensure keys
        remain unique across resamples.
      - Adjusted recipe preparation to exclude .resample_id
        consistently.
      - Refined model tuning vs non-tuning workflows for clarity and
        stability.
      - Enhanced error reporting and verbose output for improved user
        feedback.

  - Dependency updates:
    
      - tune (>= 2.0.0)
      - modeltime.resample (>= 0.3.0)

  - Version bump to 1.1.0 for CRAN submission.

                 Changes in version 1.0.5 (2025-08-28)                  

  - Development release with early updates for upcoming tune changes
    (@hfrick, #32).

                 Changes in version 1.0.4 (2024-07-19)                  

  - #31 Fixes issue with metric argument not being specified:

Error in `tune::show_best()`:
! `...` must be empty.
✖ Problematic argument:
• ..1 = metric
ℹ Did you forget to name an argument?

                 Changes in version 1.0.3 (2023-04-18)                  

  - Resubmit to CRAN (following timetk archival)

                 Changes in version 1.0.2 (2022-10-18)                  

  - Update tests for workflows mode = "regression"

                 Changes in version 1.0.1 (2022-06-09)                  

Fixes

  - Updates for hardhat 1.0.0

                 Changes in version 1.0.0 (2021-10-19)                  

NEW Nested Modeltime Ensembles

In modeltime 1.0.0, we introduced Nested Forecasting as a way to
forecast many time series iteratively. In modeltime.ensemble 1.0.0, we
introduce nested ensembles that can improve forecasting performance and
be applied to many time series iteratively. We have added:

  - ensemble_nested_average(): Apply average ensembles iteratively
  - ensemble_nested_weighted(): Apply weighted ensembles iteratively

New Vignette (Nested Ensembles)

  - Nested Ensembles

                 Changes in version 0.4.2 (2021-07-16)                  

Compatibility with modeltime 0.7.0.

  - Calibration: Added "id" feature to enable accuracy and confidence
    intervals by time series ID.

                 Changes in version 0.4.1 (2021-05-31)                  

  - Improvements for parallel processing during refitting (available in
    modeltime 0.6.0).
  - Requires modeltime 0.6.0 and parsnip 0.1.6 to align with xgboost
    upgrades.

                 Changes in version 0.4.0 (2021-04-05)                  

Recursive Ensembles

  - recursive() - The recursive() function is extended to recursive
    ensembles for both single time series and multiple time series
    models (panel data).
  - "Forecasting with Recursive Ensembles" - A new forecasting vignette
    for using recurive() with ensembles.

Fixes

  - modeltime_forecast() now returns NA when missing values are present
    in the sub-model predictions.

                 Changes in version 0.3.0 (2020-11-06)                  

Panel Data

  - Improvements made to ensemble_average(), ensemble_weighted() and
    ensemble_model_spec() to support Panel Data (i.e. when data sets
    with multiple time series groups that have possibly overlapping time
    stamps).

Changes

  - modeltime.ensemble now depends on modeltime.resample for the
    modeltime_fit_resamples() functionality.
  - modeltime_fit_resamples() moved to a new package modeltime.resample.
  - ensemble_weighted(): Now removes models that have no weight (e.g.
    loading = 0). This speeds up refitting.

                 Changes in version 0.2.0 (2020-10-09)                  

Stacked Ensembles (Breaking Changes)

The process for creating stacked ensembles is split into 2 steps:

  - Step 1: Use modeltime_fit_resamples() to generate resampled
    predictions
  - Step 2: Use ensemble_model_spec() to apply stacking using a
    model_spec

Note - modeltime_refit(stacked_ensemble) is still one step, which is the
best way to handle refitting since multiple stacked models may have
different submodel compositions. An additional argument, resamples can
be provided to train stacked ensembles made with ensemble_model_spec().

                 Changes in version 0.1.0 (2020-10-07)                  

  - Initial release of modeltime.ensemble.