Rebase (#3297)

Author: lucianopaz

RELEASE-NOTES.md

@@ -33,6 +33,7 @@
 - Rewrote `Multinomial._random` method to better handle shape broadcasting (#3271)
 - Fixed `Rice` distribution, which inconsistently mixed two parametrizations (#3286).
 - `Rice` distribution now accepts multiple parameters and observations and is usable with NUTS (#3289).
+- `sample_posterior_predictive` no longer calls `draw_values` to initialize the shape of the ppc trace. This called could lead to `ValueError`'s when sampling the ppc from a model with `Flat` or `HalfFlat` prior distributions (Fix issue #3294).
 
 
 ### Deprecations

pymc3/sampling.py

@@ -1123,17 +1123,9 @@ def sample_posterior_predictive(trace, samples=None, model=None, vars=None, size
     if progressbar:
         indices = tqdm(indices, total=samples)
 
-    varnames = [var.name for var in vars]
-
-    # draw once to inspect the shape
-    var_values = list(zip(varnames,
-                          draw_values(vars, point=model.test_point, size=size)))
     ppc_trace = defaultdict(list)
-    for varname, value in var_values:
-        ppc_trace[varname] = np.zeros((samples,) + value.shape, value.dtype)
-
     try:
-        for slc, idx in enumerate(indices):
+        for idx in indices:
             if nchain > 1:
                 chain_idx, point_idx = np.divmod(idx, len_trace)
                 param = trace._straces[chain_idx % nchain].point(point_idx)
@@ -1142,7 +1134,7 @@ def sample_posterior_predictive(trace, samples=None, model=None, vars=None, size
 
             values = draw_values(vars, point=param, size=size)
             for k, v in zip(vars, values):
-                ppc_trace[k.name][slc] = v
+                ppc_trace[k.name].append(v)
 
     except KeyboardInterrupt:
         pass
@@ -1151,7 +1143,7 @@ def sample_posterior_predictive(trace, samples=None, model=None, vars=None, size
         if progressbar:
             indices.close()
 
-    return ppc_trace
+    return {k: np.asarray(v) for k, v in ppc_trace.items()}
 
 
 def sample_ppc(*args, **kwargs):

pymc3/tests/test_sampling.py

@@ -289,6 +289,21 @@ def test_sum_normal(self):
             _, pval = stats.kstest(ppc['b'], stats.norm(scale=scale).cdf)
             assert pval > 0.001
 
+    def test_model_not_drawable_prior(self):
+        data = np.random.poisson(lam=10, size=200)
+        model = pm.Model()
+        with model:
+            mu = pm.HalfFlat('sigma')
+            pm.Poisson('foo', mu=mu, observed=data)
+            trace = pm.sample(tune=1000)
+
+        with model:
+            with pytest.raises(ValueError) as excinfo:
+                pm.sample_prior_predictive(50)
+            assert "Cannot sample" in str(excinfo.value)
+            samples = pm.sample_posterior_predictive(trace, 50)
+            assert samples['foo'].shape == (50, 200)
+
 
 class TestSamplePPCW(SeededTest):
     def test_sample_posterior_predictive_w(self):