Fix JAX implementation of Categorical

pytensor/link/jax/dispatch/random.py

@@ -169,6 +169,20 @@ def jax_sample_fn_loc_scale(op):
 
 
 @jax_sample_fn.register(ptr.BernoulliRV)
+def jax_sample_fn_bernoulli(op):
+    """JAX implementation of `BernoulliRV`."""
+
+    # We need a separate dispatch, because there is no dtype argument for Bernoulli in JAX
+    def sample_fn(rng, size, dtype, p):
+        rng_key = rng["jax_state"]
+        rng_key, sampling_key = jax.random.split(rng_key, 2)
+        sample = jax.random.bernoulli(sampling_key, p, shape=size)
+        rng["jax_state"] = rng_key
+        return (rng, sample)
+
+    return sample_fn
+
+
 @jax_sample_fn.register(ptr.CategoricalRV)
 def jax_sample_fn_no_dtype(op):
     """Generic JAX implementation of random variables."""

tests/link/jax/test_random.py

@@ -595,8 +595,16 @@ def test_random_categorical():
     g = pt.random.categorical(0.25 * np.ones(4), size=(10000, 4), rng=rng)
     g_fn = compile_random_function([], g, mode=jax_mode)
     samples = g_fn()
+    assert samples.shape == (10000, 4)
     np.testing.assert_allclose(samples.mean(axis=0), 6 / 4, 1)
 
+    # Test zero probabilities
+    g = pt.random.categorical([0, 0.5, 0, 0.5], size=(1000,), rng=rng)
+    g_fn = compile_random_function([], g, mode=jax_mode)
+    samples = g_fn()
+    assert samples.shape == (1000,)
+    assert np.all(samples % 2 == 1)
+
 
 def test_random_permutation():
     array = np.arange(4)