pymc
pymc copied to clipboard
pymc-devs
Bayesian Modeling and Probabilistic Programming in Python
## Description of your problem This issue is a spin off from a discussion in #5787. At the moment of that PR, our step methods rely on the global numpy...
## Description The internal function `pm.data.determine_coords` is called when `pm.Data(..., export_index_as_coords=True)` is invoked. Right now it can only infer `coords` from `pd.Series`, or `pd.DataFrame`, but it would be nice (and...
Related to #5646, the current `test_step_continuous` in `test_step.py` is a bit flaky and I'm not sure what it's designed to test. Can I get some insight into test so I...
When sampling a model using Compound step in serial the first chain runs with tuning, but for subsequent chains individual sampler has `tune` parameter set to `False`. The problem seems...
## When I sample larger datasets using jax (which will only sample this without error), the constant_data is not included in the inference_data output. See below. **example.** ```python coords =...
Thank you for looking at my issue. ## Description of your problem I ran the command provided on the installation guide: conda install -c conda-forge pymc3 theano-pymc mkl mkl-service conda...
### Discussed in https://github.com/pymc-devs/pymc/discussions/5579 Originally posted by **ricardoV94** March 11, 2022 I think the tests files should mirror the original codebase. So they should be in a file with the...
A recent PR #5298 added GaussianRandomWalk distribution to v4. It would be possible to use AePPL to obtain the Aesara graph corresponding to desired `rv_op` and retrieve the correct log-probability...
In discussion with @OriolAbril, it was brought up that, for model comparison, it is useful to have access to the probability of each observation/time step. Currently, the GaussianRandomWalk collapses these...
No reason to require a double click in these cases... unless it's too much work to change the behavior of course :) 