Olivier Sprangers
Olivier Sprangers
- Adds possibility to use historical exogenous variables. Basically, historical exogenous variables are those variables in `df` that are not in `X_df` - Updates docs to show how historical exogenous...
Fixes some docs errors: - Mintlify doesn't display text if the text is written in the same markdown cell as the header. Thus, this PR splits off all headers +...
This reduces the minimum required size for finetuning, by reducing the values in the checks.
### What happened + What you expected to happen Cross-validating + finetuning with Polars on a Float32 target array fails. The following code fails, whereas removing the line that casts...
### Description [Mintlify](https://nixtlaverse.nixtla.io/nixtla/docs/getting-started/introduction.html) doesn't show our pricing currently. I don't yet know what's going on, it's in the sidebar in mint.json, maybe we just need to re-run the documentation workflows....
Deprecates the old evaluation method in: - [x] ~~Tests~~ Skipped because not all lossses have been transferred to utilsforecast yet. - [x] ~~Experiments~~ Skipped because not all lossses have been...
- Adds type hints to probabilistic methods - Adds test for sparse S, which uncovered a number of bugs in applying probabilistic methods to sparse S matrices.
### What happened + What you expected to happen Probalistic methods are generally typed with Numpy arrays, however they also work with Sparse methods. These methods however generate scipy sparse...
### What happened + What you expected to happen CV with refit=True on Pandas input DataFrame fails because there seems to be a mismatch in associated dates when the prediciton...
Adds step_size to conformal prediction intervals. Need to think about what this means for theoretical purposes, and if the min_samples is correct.