Allow for batched alpha in StickBreakingWeights

pymc/distributions/multivariate.py

@@ -2192,32 +2192,25 @@ def make_node(self, rng, size, dtype, alpha, K):
         alpha = at.as_tensor_variable(alpha)
         K = at.as_tensor_variable(intX(K))
 
-        if alpha.ndim > 0:
-            raise ValueError("The concentration parameter needs to be a scalar.")
-
         if K.ndim > 0:
             raise ValueError("K must be a scalar.")
 
         return super().make_node(rng, size, dtype, alpha, K)
 
-    def _infer_shape(self, size, dist_params, param_shapes=None):
-        alpha, K = dist_params
-
-        size = tuple(size)
-
-        return size + (K + 1,)
+    def _supp_shape_from_params(self, dist_params, **kwargs):
+        K = dist_params[1]
+        return (K + 1,)
 
     @classmethod
     def rng_fn(cls, rng, alpha, K, size):
         if K < 0:
             raise ValueError("K needs to be positive.")
 
         if size is None:
-            size = (K,)
-        elif isinstance(size, int):
-            size = (size,) + (K,)
+            size = alpha.shape + (K,)
+            alpha = alpha[..., np.newaxis]
         else:
-            size = tuple(size) + (K,)
+            size = size + (K,)
 
         betas = rng.beta(1, alpha, size=size)
 

pymc/tests/test_distributions.py

@@ -24,6 +24,7 @@
 from aesara.tensor.random.utils import broadcast_params
 
 from pymc.distributions.continuous import get_tau_sigma
+from pymc.distributions.dist_math import betaln
 from pymc.util import UNSET
 
 try:
@@ -952,6 +953,38 @@ def test_hierarchical_obs_logp():
     assert not any(isinstance(o, RandomVariable) for o in ops)
 
 
+def _stickbreakingweights_logpdf(value, alpha, K):
+    logp = -at.sum(
+        at.log(
+            at.cumsum(
+                value[..., ::-1],
+                axis=-1,
+            )
+        ),
+        axis=-1,
+    )
+    logp += -K * betaln(1, alpha)
+    logp += alpha * at.log(value[..., -1])
+    logp = at.switch(
+        at.or_(
+            at.any(
+                at.and_(at.le(value, 0), at.ge(value, 1)),
+                axis=-1,
+            ),
+            at.or_(
+                at.bitwise_not(at.allclose(value.sum(-1), 1)),
+                at.neq(value.shape[-1], K + 1),
+            ),
+        ),
+        -np.inf,
+        logp,
+    )
+    return logp.eval()
+
+
+stickbreakingweights_logpdf = np.vectorize(_stickbreakingweights_logpdf, signature="(n),(),()->()")
+
+
 class TestMatchesScipy:
     def test_uniform(self):
         check_logp(
@@ -2312,6 +2345,28 @@ def test_stickbreakingweights_invalid(self):
         assert pm.logp(sbw, np.array([0.4, 0.3, 0.2, -0.1])).eval() == -np.inf
         assert pm.logp(sbw_wrong_K, np.array([0.4, 0.3, 0.2, 0.1])).eval() == -np.inf
 
+    @pytest.mark.parametrize(
+        "value, alpha, K",
+        [
+            (np.array([5, 4, 3, 2, 1]) / 15, [0.5, 1.0, 2.0], 19),
+            (
+                np.append(0.5 ** np.arange(1, 20), 0.5**20),
+                np.arange(1, 7, dtype="float64").reshape(2, 3),
+                4,
+            ),
+        ],
+    )
+    def test_stickbreakingweights_vectorized(self, value, alpha, K):
+        with Model():
+            sbw = StickBreakingWeights("sbw", alpha=alpha, K=K, transform=None)
+        pt = {"sbw": value}
+        assert_almost_equal(
+            pm.logp(sbw, value).eval(),
+            stickbreakingweights_logpdf(value, alpha, K),
+            decimal=select_by_precision(float64=6, float32=2),
+            err_msg=str(pt),
+        )
+
     @aesara.config.change_flags(compute_test_value="raise")
     def test_categorical_bounds(self):
         with Model():

pymc/tests/test_distributions_random.py

@@ -1321,6 +1321,18 @@ def check_basic_properties(self):
         assert np.all(draws <= 1)
 
 
+class TestStickBreakingWeights_1D_alpha(BaseTestDistributionRandom):
+    pymc_dist = pm.StickBreakingWeights
+    pymc_dist_params = {"alpha": [1.0, 2.0, 3.0], "K": 19}
+    expected_rv_op_params = {"alpha": [1.0, 2.0, 3.0], "K": 19}
+    sizes_to_check = [None]
+    sizes_expected = [(3, 20)]
+    checks_to_run = [
+        "check_pymc_params_match_rv_op",
+        "check_rv_size",
+    ]
+
+
 class TestCategorical(BaseTestDistributionRandom):
     pymc_dist = pm.Categorical
     pymc_dist_params = {"p": np.array([0.28, 0.62, 0.10])}