pymc icon indicating copy to clipboard operation
pymc copied to clipboard

Bayesian Modeling and Probabilistic Programming in Python

Results 537 pymc issues
Sort by recently updated
recently updated
newest added

### Issue with current documentation: [`pymc.sample`](https://www.pymc.io/projects/docs/en/v5.10.0/api/generated/pymc.sample.html#pymc-sample) was replaced by [`pymc.sampling.mcmc.sample`](https://www.pymc.io/projects/docs/en/latest/api/generated/pymc.sampling.mcmc.sample.html#pymc-sampling-mcmc-sample) in the documentation after `v.5.10.0`, but `pymc.sample` can still be called under `pymc-5.10.2`. ### Idea or request for content: _No...

docs

## Description FutureWarning when users access sumto1 transform to let them know it's deprecated and will be removed in the future, and a test that verified the warning is working...

maintenance
major

### Description Because we don't save transformed variables in the returned InferenceData (why not?) it's not easy to evaluate the model logp once we have a trace. ~~One could rewrite...

enhancements
help wanted
trace-backend

Squeezing stats by default removes the `"stepmethod"` dimension (which we don't name explicitly), but this does not always work [for `object`-dtyped stats](https://github.com/pymc-devs/pymc/blob/d47dac0e92da2398206b42d3748bc63ff7ff7103/pymc/backends/base.py#L210-L215). This is (very often) the case for the...

shape problem
trace-backend
major

### Description Sometimes models with large Deterministics run out of RAM during sampling. One could compute the Deterministics afterwards easily, but would need to create the same model twice, once...

enhancements
request discussion

### Describe the issue: So I am trying to implement a MvNormal where the covariance itself is hierarchically dependent on other priors. I have also sort of described the issue...

bug
shape problem
VI

### Describe the issue: I wanted to test the detection for several breaking points in a time series, so I came up with the following model (cf. code example) to...

bug
samplers

### Describe the issue: In `pymc.sampling.jax.sample_numpyro_nuts`, a jitter is always applied to the initial values through function `_get_batched_jittered_initial_points()`: https://github.com/pymc-devs/pymc/blob/904a0eaac216732bc358dd91680cd428d95704f0/pymc/sampling/jax.py#L662-L668 This function actually accepts a `jitter` argument that allows jittering to...

bug

This PR will add implementation of default rewrites for nodes with inner `fgraph` and related with issue #6697. For now I just added new test cases ---- :books: Documentation preview...

enhancements

Closes #6920 ---- :books: Documentation preview :books:: https://pymc--6997.org.readthedocs.build/en/6997/

enhancements
GP