pymc

pymc copied to clipboard

Dims is not respected when broadcasting parameters

10

The dims argument is not respected when a random variable is created with parameters that broadcast to the desired output dimensions: ```python coords = { "dim1": pd.RangeIndex(10), "dim2": pd.RangeIndex(7) }...

aseyboldt

**What is this PR about?** ... Description: Change Model docstring Issue name Model docstring is outdated Issue number: #6040 **Checklist** + [x ] Explain important implementation details 👆 + [x...

alekracicot

Added networkx export functionality

12

Apologies for delay but finally got around to updating the latest version of model_graph.py to include networkx export function as discussed in [5677](https://github.com/pymc-devs/pymc/pull/5677) This creates a method for exporting the...

jonititan

Variables with shared inputs are always resampled from the prior in sample_posterior_predictive

14

## Description of your problem Using the model.add_coord method appears to break pm.sample_posterior_predictive. Here’s a simple example, estimating the loc and scale of three independent normals: **Please provide a minimal,...

jcstucki

Data objects not connected to binomial, Poisson likelihood when parameters have non-scalar shape

2

## Description of your problem When using `Data` or `MutableData` objects with binomial or Poisson likelihoods, they are not properly added to the model graph when the likelihood has non-scalar...

fonnesbeck

Model docstring is outdated

2

We no longer have self.Var("name", dist) to register random variables. We need to fix that with a correct API call https://github.com/pymc-devs/pymc/blob/ad16bf48a1703b80cee3d3fef9df1dd0ae552d7c/pymc/model.py#L469

ferrine

Allow for batched `alpha` in `StickBreakingWeights`

1

**What is this PR about?** Addressing #5383 This enables `StickBreakingWeight`'s `alpha` to accept batched data (>2D), make the `infer_shape` work with batched data, and fix the `rng_fn` by broadcasting alpha...

purna135

Updated pymc.simulator docstring (typos, defaults, type description)

9

**What is this PR about?** fixing a number of issues with the pymc.simulator docstring. ... **Checklist** + [ ] Explain important implementation details 👆 + [ ] Make sure that...

daniel-saunders-phil

Prediction could be better if shape of observed would not need to be considered

13

When I have a simple model: ```python import pymc as pm with pm.Model() as m: x = pm.MutableData("x", input_data) y = pm.MutableData("y" , output_data) beta = pm.Normal("beta") y_est = x...

twiecki

Fix JAX sampling funcs overwriting existing var's dims and coords

1

**What is this PR about?** The original solution to fix Issue #5932 overwrites existing dimensions and coordinate data on the variables when creating the ArviZ InferenceData object. (See the issue...

jhrcook