Remove Dirichlet distribution type restrictions (#4000)

* Remove Dirichlet distribution type restrictions

Closes #3999.

* Add missing Dirichlet shape parameters to tests

* Remove Dirichlet positive concentration parameter constructor tests

This test can't be performed in the constructor if we're allowing Theano-type
distribution parameters.

* Add a hack to statically infer Dirichlet argument shapes

Co-authored-by: Brandon T. Willard <[email protected]>

Author: brandonwillard

pymc3/distributions/multivariate.py

@@ -23,6 +23,7 @@
 
 from scipy import stats, linalg
 
+from theano.gof.op import get_test_value
 from theano.tensor.nlinalg import det, matrix_inverse, trace, eigh
 from theano.tensor.slinalg import Cholesky
 import pymc3 as pm
@@ -487,22 +488,23 @@ class Dirichlet(Continuous):
     def __init__(self, a, transform=transforms.stick_breaking,
                  *args, **kwargs):
 
-        if not isinstance(a, pm.model.TensorVariable):
-            if not isinstance(a, list) and not isinstance(a, np.ndarray):
-                raise TypeError(
-                    'The vector of concentration parameters (a) must be a python list '
-                    'or numpy array.')
-            a = np.array(a)
-            if (a <= 0).any():
-               raise ValueError("All concentration parameters (a) must be > 0.")
-
-        shape = np.atleast_1d(a.shape)[-1]
+        if kwargs.get('shape') is None:
+            warnings.warn(
+                (
+                    "Shape not explicitly set. "
+                    "Please, set the value using the `shape` keyword argument. "
+                    "Using the test value to infer the shape."
+                ),
+                DeprecationWarning
+            )
+            try:
+                kwargs['shape'] = get_test_value(tt.shape(a))
+            except AttributeError:
+                pass
 
-        kwargs.setdefault("shape", shape)
         super().__init__(transform=transform, *args, **kwargs)
 
         self.size_prefix = tuple(self.shape[:-1])
-        self.k = tt.as_tensor_variable(shape)
         self.a = a = tt.as_tensor_variable(a)
         self.mean = a / tt.sum(a)
 
@@ -569,14 +571,13 @@ def logp(self, value):
         -------
         TensorVariable
         """
-        k = self.k
         a = self.a
 
         # only defined for sum(value) == 1
         return bound(tt.sum(logpow(value, a - 1) - gammaln(a), axis=-1)
                      + gammaln(tt.sum(a, axis=-1)),
                      tt.all(value >= 0), tt.all(value <= 1),
-                     k > 1, tt.all(a > 0),
+                     np.logical_not(a.broadcastable), tt.all(a > 0),
                      broadcast_conditions=False)
 
     def _repr_latex_(self, name=None, dist=None):

pymc3/tests/test_dist_math.py

@@ -126,11 +126,11 @@ def test_multinomial_bound():
     n = x.sum()
 
     with pm.Model() as modelA:
-        p_a = pm.Dirichlet('p', floatX(np.ones(2)))
+        p_a = pm.Dirichlet('p', floatX(np.ones(2)), shape=(2,))
         MultinomialA('x', n, p_a, observed=x)
 
     with pm.Model() as modelB:
-        p_b = pm.Dirichlet('p', floatX(np.ones(2)))
+        p_b = pm.Dirichlet('p', floatX(np.ones(2)), shape=(2,))
         MultinomialB('x', n, p_b, observed=x)
 
     assert np.isclose(modelA.logp({'p_stickbreaking__': [0]}),

pymc3/tests/test_distributions.py

@@ -1328,17 +1328,14 @@ def test_dirichlet(self, n):
             Dirichlet, Simplex(n), {"a": Vector(Rplus, n)}, dirichlet_logpdf
         )
 
-    @pytest.mark.parametrize("n", [3, 4])
-    def test_dirichlet_init_fail(self, n):
-        with Model():
-            with pytest.raises(
-                ValueError, match=r"All concentration parameters \(a\) must be > 0."
-            ):
-                _ = Dirichlet("x", a=np.zeros(n), shape=n)
-            with pytest.raises(
-                ValueError, match=r"All concentration parameters \(a\) must be > 0."
-            ):
-                _ = Dirichlet("x", a=np.array([-1.0] * n), shape=n)
+    def test_dirichlet_shape(self):
+        a = tt.as_tensor_variable(np.r_[1, 2])
+        with pytest.warns(DeprecationWarning):
+            dir_rv = Dirichlet.dist(a)
+            assert dir_rv.shape == (2,)
+
+        with pytest.warns(DeprecationWarning), theano.change_flags(compute_test_value="ignore"):
+            dir_rv = Dirichlet.dist(tt.vector())
 
     def test_dirichlet_2D(self):
         self.pymc3_matches_scipy(

pymc3/tests/test_distributions_random.py

@@ -912,15 +912,15 @@ def test_mixture_random_shape():
                         nr.poisson(9, size=10)])
     with pm.Model() as m:
         comp0 = pm.Poisson.dist(mu=np.ones(2))
-        w0 = pm.Dirichlet('w0', a=np.ones(2))
+        w0 = pm.Dirichlet('w0', a=np.ones(2), shape=(2,))
         like0 = pm.Mixture('like0',
                            w=w0,
                            comp_dists=comp0,
                            observed=y)
 
         comp1 = pm.Poisson.dist(mu=np.ones((20, 2)),
                                 shape=(20, 2))
-        w1 = pm.Dirichlet('w1', a=np.ones(2))
+        w1 = pm.Dirichlet('w1', a=np.ones(2), shape=(2,))
         like1 = pm.Mixture('like1',
                            w=w1,
                            comp_dists=comp1,
@@ -967,15 +967,15 @@ def test_mixture_random_shape_fast():
                         nr.poisson(9, size=10)])
     with pm.Model() as m:
         comp0 = pm.Poisson.dist(mu=np.ones(2))
-        w0 = pm.Dirichlet('w0', a=np.ones(2))
+        w0 = pm.Dirichlet('w0', a=np.ones(2), shape=(2,))
         like0 = pm.Mixture('like0',
                            w=w0,
                            comp_dists=comp0,
                            observed=y)
 
         comp1 = pm.Poisson.dist(mu=np.ones((20, 2)),
                                 shape=(20, 2))
-        w1 = pm.Dirichlet('w1', a=np.ones(2))
+        w1 = pm.Dirichlet('w1', a=np.ones(2), shape=(2,))
         like1 = pm.Mixture('like1',
                            w=w1,
                            comp_dists=comp1,

pymc3/tests/test_mixture.py

@@ -79,7 +79,7 @@ def test_dimensions(self):
 
     def test_mixture_list_of_normals(self):
         with Model() as model:
-            w = Dirichlet('w', floatX(np.ones_like(self.norm_w)))
+            w = Dirichlet('w', floatX(np.ones_like(self.norm_w)), shape=self.norm_w.size)
             mu = Normal('mu', 0., 10., shape=self.norm_w.size)
             tau = Gamma('tau', 1., 1., shape=self.norm_w.size)
             Mixture('x_obs', w,
@@ -98,7 +98,7 @@ def test_mixture_list_of_normals(self):
 
     def test_normal_mixture(self):
         with Model() as model:
-            w = Dirichlet('w', floatX(np.ones_like(self.norm_w)))
+            w = Dirichlet('w', floatX(np.ones_like(self.norm_w)), shape=self.norm_w.size)
             mu = Normal('mu', 0., 10., shape=self.norm_w.size)
             tau = Gamma('tau', 1., 1., shape=self.norm_w.size)
             NormalMixture('x_obs', w, mu, tau=tau, observed=self.norm_x)
@@ -135,7 +135,7 @@ def test_normal_mixture_nd(self, nd, ncomp):
         with Model() as model0:
             mus = Normal('mus', shape=comp_shape)
             taus = Gamma('taus', alpha=1, beta=1, shape=comp_shape)
-            ws = Dirichlet('ws', np.ones(ncomp))
+            ws = Dirichlet('ws', np.ones(ncomp), shape=(ncomp,))
             mixture0 = NormalMixture('m', w=ws, mu=mus, tau=taus, shape=nd,
                                      comp_shape=comp_shape)
             obs0 = NormalMixture('obs', w=ws, mu=mus, tau=taus, shape=nd,
@@ -145,7 +145,7 @@ def test_normal_mixture_nd(self, nd, ncomp):
         with Model() as model1:
             mus = Normal('mus', shape=comp_shape)
             taus = Gamma('taus', alpha=1, beta=1, shape=comp_shape)
-            ws = Dirichlet('ws', np.ones(ncomp))
+            ws = Dirichlet('ws', np.ones(ncomp), shape=(ncomp,))
             comp_dist = [Normal.dist(mu=mus[..., i], tau=taus[..., i],
                                      shape=nd)
                          for i in range(ncomp)]
@@ -163,7 +163,7 @@ def test_normal_mixture_nd(self, nd, ncomp):
             # comp_dists.
             mus = Normal('mus', shape=comp_shape)
             taus = Gamma('taus', alpha=1, beta=1, shape=comp_shape)
-            ws = Dirichlet('ws', np.ones(ncomp))
+            ws = Dirichlet('ws', np.ones(ncomp), shape=(ncomp,))
             if len(nd) > 1:
                 if nd[-1] != ncomp:
                     with pytest.raises(ValueError):
@@ -208,7 +208,7 @@ def test_normal_mixture_nd(self, nd, ncomp):
 
     def test_poisson_mixture(self):
         with Model() as model:
-            w = Dirichlet('w', floatX(np.ones_like(self.pois_w)))
+            w = Dirichlet('w', floatX(np.ones_like(self.pois_w)), shape=self.pois_w.shape)
             mu = Gamma('mu', 1., 1., shape=self.pois_w.size)
             Mixture('x_obs', w, Poisson.dist(mu), observed=self.pois_x)
             step = Metropolis()
@@ -224,7 +224,7 @@ def test_poisson_mixture(self):
 
     def test_mixture_list_of_poissons(self):
         with Model() as model:
-            w = Dirichlet('w', floatX(np.ones_like(self.pois_w)))
+            w = Dirichlet('w', floatX(np.ones_like(self.pois_w)), shape=self.pois_w.shape)
             mu = Gamma('mu', 1., 1., shape=self.pois_w.size)
             Mixture('x_obs', w,
                     [Poisson.dist(mu[0]), Poisson.dist(mu[1])],
@@ -247,7 +247,7 @@ def test_mixture_of_mvn(self):
         cov2 = np.diag([2.5, 3.5])
         obs = np.asarray([[.5, .5], mu1, mu2])
         with Model() as model:
-            w = Dirichlet('w', floatX(np.ones(2)), transform=None)
+            w = Dirichlet('w', floatX(np.ones(2)), transform=None, shape=(2,))
             mvncomp1 = MvNormal.dist(mu=mu1, cov=cov1)
             mvncomp2 = MvNormal.dist(mu=mu2, cov=cov2)
             y = Mixture('x_obs', w, [mvncomp1, mvncomp2],
@@ -291,13 +291,13 @@ def test_mixture_of_mixture(self):
                 sigma=1,
                 shape=nbr)
             # weight vector for the mixtures
-            g_w = Dirichlet('g_w', a=floatX(np.ones(nbr)*0.0000001), transform=None)
-            l_w = Dirichlet('l_w', a=floatX(np.ones(nbr)*0.0000001), transform=None)
+            g_w = Dirichlet('g_w', a=floatX(np.ones(nbr)*0.0000001), transform=None, shape=(nbr,))
+            l_w = Dirichlet('l_w', a=floatX(np.ones(nbr)*0.0000001), transform=None, shape=(nbr,))
             # mixture components
             g_mix = Mixture.dist(w=g_w, comp_dists=g_comp)
             l_mix = Mixture.dist(w=l_w, comp_dists=l_comp)
             # mixture of mixtures
-            mix_w = Dirichlet('mix_w', a=floatX(np.ones(2)), transform=None)
+            mix_w = Dirichlet('mix_w', a=floatX(np.ones(2)), transform=None, shape=(2,))
             mix = Mixture('mix', w=mix_w,
                           comp_dists=[g_mix, l_mix],
                           observed=np.exp(self.norm_x))
@@ -378,7 +378,7 @@ def build_toy_dataset(N, K):
         X, y = build_toy_dataset(N, K)
 
         with pm.Model() as model:
-            pi = pm.Dirichlet('pi', np.ones(K))
+            pi = pm.Dirichlet('pi', np.ones(K), shape=(K,))
 
             comp_dist = []
             mu = []