Jan

icon indicating a website link janfb.github.io

icon company @aai-institute icon location Munich icon location Researcher @aai-institute. Working on Bayesian Machine Learning and Simulation Based Inference

Results 241 comments of Jan

prepare_for_sbi causes import error with the latest scipy

With `arviz=0.18.0` being on pypi, a fresh installation of `sbi` 0.22.0 will install `scipy` 1.13.1 and `arviz` 0.18.0, which should fix the problem. Please note that in the upcoming release...

MCMC sampler generates `inf` values

Thanks for creating this! Relevant for #910

Make code formatting easier and less painful

Update: we pinned `pre-commit=3.5.0` and require `ruff>=0.3.3`. For the moment this results in consistent code formatting it seems.

Make code formatting easier and less painful

closing this for now.

base_dist methods for nflows and zuko

Hi @JuliaLinhart, I am not sure what is meant here. Can you please give more context?

Provide a devcontainer to be used in Codespace

related to #1056

Refactor abstract classes for custom density estimators

will be relevant for #963 and the `RatioEstimator` as well @bkmi

Refactor abstract classes for custom density estimators

This is solved now: We have `ConditionalEstimator` abstract base class in `neural_nets/density_estimators` that takes care of shapes and requires children to implement `loss`, `log_prob` and `sample`. `ConditionalDensityEstimator` is the class...

StandardizeInputs requires a list of exactly two elements

closed with #1097

ENH add h npe

Is this ready for review?