Ari Hartikainen

Update pystan in requirements

9

Currently latest pystan is 3.0.0 which is incompatible with 2.x. https://pypi.org/project/pystan Following PEP 440 could be used https://www.python.org/dev/peps/pep-0440/#compatible-release pystan~=2.14

Bayesian variable (one-object mode)

8

I think we could wrap our mcmc parameters in a custom class that would make it easier to work with the mcmc results (in the spirit of rval (?) in...

**Describe the bug** CI crashes with latest numpy for pyjags https://dev.azure.com/ArviZ/ArviZ/_build/results?buildId=5398&view=logs&j=7f538728-4589-5563-0e87-d0823eb40420&t=bc0ad562-2d38-5af9-a6a9-06323c6c7eeb @michaelnowotny ``` 2022-01-03T20:09:03.6277524Z ##[section]Starting: pytest 2022-01-03T20:09:03.6283298Z ============================================================================== 2022-01-03T20:09:03.6283814Z Task : Command line 2022-01-03T20:09:03.6284220Z Description : Run a command line...

Check exceptions and exception messages

5

Check exceptions and exception messages reasonable and intuitive. We have a few locations with (only) Exception-class raised and a lot of ValueErrors that might better be TypeErrors or something similar....

R and Julia

5

We should have some functionality to transform results gathered with R or Julia to our InferenceData format (netCDF4). These could be native (written in R and Julia) modules.

Hi, quick question if we had some restrictions for the Rdump format? For example, R dataframes cannot be read with PyStan functions. https://github.com/stan-dev/example-models/blob/master/misc/gaussian-process/gp-fit-ARD.data.R If they are "valid", then we need...

Save InferenceData attrs

5

closes #2109 ## Description This will save main level attrs to "__netcdf4_attrs" and also unload if present. ## Checklist - [ ] Follows [official](https://github.com/arviz-devs/arviz/blob/main/CONTRIBUTING.md#pull-request-checklist) PR format - [ ] Includes...

Bokeh, transform all colors to hex

24

Input all colors to bokeh functions as hex. We can use matplotlib helper functions to transform special syntax color="C0" etc

VI - Variational Inference support

9

To support VI we need to decide what different aspect we should implement. ## Data structure We probably need a special data structure for VI results. It could contain: -...

fix: Update `num_samples_saved` formula in `fit.py`

3

Fixes #366 Fixes the formula to calculate `num_samples_saved`. Updates a test in `tests/test_normal.py` to hit this specific problem for both `num_samples` and `num_warmup` when `save_warmup=True`.