Rohan George Thampi comments

Results 17 comments of


                                            Rohan George Thampi

[ENHANCEMENT] Make fitting with exogenous variables more relaible

Any update on this? I was looking for a way to use external regressors in hierarchical Prophet models

Spot instance support

Any update on this?

[Feature Request] Smarter variable type inference

Hmm validation could be done based on a random sample or to be more thorough scan all values and using a confidence threshold although the latter might be computationally expensive

[ENH] Add adapter for `arch` package.

Any update on this enhancement?

Feature Request: CLI Support & Lossless Resize

Nope a command line interface for easier batch conversion and lossless resizing options

[Question] Early stopping implementation

Yes it has already been implemented for HistGradientBoostingClassifier (https://github.com/scikit-learn/scikit-learn/issues/14303) although they're using `train_test_split`

Add customizability to how features are stacked

@kmax12 I'm aware that they're two different features. I want to only keep the Aggregation(Transform) features since they are more meaningful. Yes I've seen the `drop_contains` and `drop_exact` parameters but...

Simplify variable type and custom primitive parameterization

@rwedge what about custom primitives? Can they be passed as snake case strings?

Simplify variable type and custom primitive parameterization

Any update on this?

Feature naming options

I've updated the comment. Mainly to add some customizability and improve readability of the column names