Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

[ENH] All univariate _SeriesToSeriesTransformer should internally use the ColumnwiseTransformer to make them multivariate

See original GitHub issue

Is your feature request related to a problem? Please describe. All univariate _SeriesToSeriesTransformer should internally use the ColumnwiseTransformer to make them multivariate. This should however be mentioned in the docstring that it internally applies just a ColumnwiseTransformer and that the algorithm itself is still a univariate.

Additional context Refactor Detrender which internally uses sth similar to ColumnwiseTransformer which should be raplaced with a ColumnwiseTransformer then also.

tagging @SveaMeyer13 who wrote the ColumnwiseTransformer

Issue Analytics

State:
Created 2 years ago
Reactions:1
Comments:11

Top GitHub Comments

2reactions

SveaMeyer13commented, Sep 6, 2021

Happy to work on this 😃

1reaction

mloningcommented, Sep 10, 2021

I think we should use the column-ensembling as the default fall-back option for transformers and implement that in the base class (still not sure about forecasters). Concrete transformers should have the option to override that option if they implement bespoke ways of handling both univariate and multivariate data by using the combined pd.Series/pd.DataFrame interface. I think that’s what the tags would allow us to do quite easily.

Top Results From Across the Web

[ENH] All univariate _SeriesToSeriesTransformer should internally ...

All univariate _SeriesToSeriesTransformer should internally use the ColumnwiseTransformer to make them multivariate. This should however be mentioned in the ...

ColumnwiseTransformer — sktime documentation

Returns an inverse-transformed version of Z by iterating over specified columns and applying the univariate series transformer to them. Only works if self....

Learning Graph Structures with Transformer for Multivariate ...

This paper presented GTA, a new framework for multivariate time series anomaly detection that involves automatically learning a graph structure, graph ...

A Fusion Transformer for Multivariable Time Series Forecasting

Instead, it analyzes the complete sequence of input and use self-attention mechanisms [17] to learn the sequence's temporal relationships, potentially making it ......

Univariate vs Multivariate Time Series Forecasting with ...

A comparative study shows that our setting outperforms state-of-the-art Transformers in the multivariate setting in benchmark datasets.