Skip to content

feat: add quantiles support for missing models#336

Merged
AzulGarza merged 3 commits intomainfrom
feat/pred-intervals
Apr 25, 2026
Merged

feat: add quantiles support for missing models#336
AzulGarza merged 3 commits intomainfrom
feat/pred-intervals

Conversation

@AzulGarza
Copy link
Copy Markdown
Member

this pr adds quantiles to missing models

Copilot AI review requested due to automatic review settings April 25, 2026 19:28
Copy link
Copy Markdown
Contributor

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Adds quantile-forecast support to previously non-probabilistic wrappers (AutoLGBM, AutoNHITS, AutoTFT) so they can return consistent model-q-{percentile} columns across the library.

Changes:

  • Enable quantile outputs for NeuralForecast AutoNHITS/AutoTFT via MQLoss and post-processing through QuantileConverter.
  • Enable quantile outputs for MLForecast AutoLGBM via conformal prediction intervals (PredictionIntervals) and conversion to quantile columns.
  • Update model tests and add a v0.0.26 changelog entry describing the new quantile support and limitations.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated no comments.

File Description
timecopilot/models/neural.py Adds quantile support for AutoNHITS/AutoTFT by training with MQLoss when quantiles are requested and converting outputs to -q- columns.
timecopilot/models/ml.py Adds quantile support for AutoLGBM using conformal prediction intervals and QuantileConverter to emit -q- columns.
tests/models/test_models.py Updates quantile/level behavior expectations for AutoLGBM/AutoNHITS/AutoTFT and adjusts monotonicity assertions accordingly.
docs/changelogs/v0.0.26.md Documents the new quantile support and usage example for the affected models.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

@AzulGarza AzulGarza merged commit 63059d9 into main Apr 25, 2026
9 of 10 checks passed
@AzulGarza AzulGarza deleted the feat/pred-intervals branch April 25, 2026 20:35
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants