Updated release notes with changes in this PR. Added a normalisation …

…to the invalid p parameter in the failing test_multivariate_observations.

RELEASE-NOTES.md

@@ -132,6 +132,7 @@ This includes API changes we did not warn about since at least `3.11.0` (2021-01
 - `math.log1mexp` and `math.log1mexp_numpy` will expect negative inputs in the future. A `FutureWarning` is now raised unless `negative_input=True` is set (see [#4860](https://github.com/pymc-devs/pymc/pull/4860)).
 - Changed name of `Lognormal` distribution to `LogNormal` to harmonize CamelCase usage for distribution names.
 - Attempt to iterate over MultiTrace will raise NotImplementedError.
+- Removed silent normalisation of p-values passed to Categorical and Multinomial distribution. Instead, a UserWarning is raised when p-values do not sum to 1.0 and then normalisation is done. Normalisation in Categorical now takes place in `dist` rather than `logp`. In addition, negative p-values now raise ValueError.
 - ...
 
 

pymc/tests/test_idata_conversion.py

@@ -588,6 +588,7 @@ def test_multivariate_observations(self):
         data = np.random.multinomial(20, [0.2, 0.3, 0.5], size=20)
         with pm.Model(coords=coords):
             p = pm.Beta("p", 1, 1, size=(3,))
+            p = p / p.sum()
             pm.Multinomial("y", 20, p, dims=("experiment", "direction"), observed=data)
             idata = pm.sample(draws=50, chains=2, tune=100, return_inferencedata=True)
         test_dict = {