Jacob
Jacob
For example, right now "plot.time_bin_analysis" has an argument for an lme4 growth-curve model. This is too specialized.
- "Preparing data" could be more concise. - "Window analysis" should be much more concise. Potentially also warn that transformations do not make data normal in the example data. -...
To avoid cluttering up issues forum, will put brainstorming for all future enhancements here.
Would be great to add a few standard enhancements to the setup: - Cacheing: It would be good to cache certain steps. For example, currently we re-install pytorch from scratch...
One notable issue is usage of `kwargs`. For example, `StateSpaceModel().fit()` ultimately calls `StateSpaceModel().forward()`, and passing `kwargs` on to that method. This makes it difficult for someone who's working with `fit()`...
Something like https://github.com/uber/orbit/blob/dev/examples/backtest.ipynb