Add random kwarg to DensityDist #2106

RELEASE-NOTES.md

@@ -7,6 +7,8 @@
 - Add `logit_p` keyword to `pm.Bernoulli`, so that users can specify the logit
   of the success probability. This is faster and more stable than using
   `p=tt.nnet.sigmoid(logit_p)`.
+- Add `random` keyword to `pm.DensityDist` thus enabling users to pass custom random method
+  which in turn makes sampling from a `DensityDist` possible. 
 
 ### Deprecations
 

pymc3/distributions/distribution.py

@@ -178,12 +178,21 @@ def __init__(self, shape=(), dtype=None, defaults=('median', 'mean', 'mode'),
 class DensityDist(Distribution):
     """Distribution based on a given log density function."""
 
-    def __init__(self, logp, shape=(), dtype=None, testval=0, *args, **kwargs):
+    def __init__(self, logp, shape=(), dtype=None, testval=0, random=None, *args, **kwargs):
         if dtype is None:
             dtype = theano.config.floatX
         super(DensityDist, self).__init__(
             shape, dtype, testval, *args, **kwargs)
         self.logp = logp
+        self.rand = random
+
+    def random(self, *args, **kwargs):
+        if self.rand is not None:
+            return self.rand(*args, **kwargs)
+        else:
+            raise ValueError("Distribution was not passed any random method "
+                            "Define a custom random method and pass it as kwarg random")
+
 
 
 def draw_values(params, point=None):

pymc3/tests/test_distributions_random.py

@@ -688,3 +688,47 @@ def ref_rand(size, w, mu, sd):
                      'sd': Domain([[1.5, 2., 3.]], edges=(None, None))}, 
                      size=1000,
                      ref_rand=ref_rand)
+
+    def test_density_dist(self):
+        def ref_rand(size, mu, sd):
+            return st.norm.rvs(size=size, loc=mu, scale=sd)
+
+        class TestDensityDist(pm.DensityDist):
+
+            def __init__(self, **kwargs):
+                norm_dist = pm.Normal.dist()
+                super(TestDensityDist, self).__init__(logp=norm_dist.logp, random=norm_dist.random)
+
+        pymc3_random(TestDensityDist, {},ref_rand=ref_rand)
+
+        def check_model_samplability(self):
+            model = pm.Model()
+            with model:
+                normal_dist = pm.Normal.dist()
+                density_dist = pm.DensityDist('density_dist', normal_dist.logp, random=normal_dist.random)
+                step = pm.Metropolis()
+                trace = pm.sample(100, step, tuning=0)
+
+            try:
+                ppc = pm.sample_ppc(trace, samples=500, model=model, size=100)
+                if len(ppc) == 0:
+                    npt.assert_true(len(ppc) == 0, 'length of ppc sample is zero')
+            except:
+                assert False
+
+        def check_scipy_distributions(self):
+            model = pm.Model()
+            with model:
+                norm_dist_logp = st.norm.logpdf
+                norm_dist_random = np.random.normal
+                density_dist = pm.DensityDist('density_dist', normal_dist_logp, random=normal_dist_random)
+                step = pm.Metropolis()
+                trace = pm.sample(100, step, tuning=0)
+
+            try:
+                ppc = pm.sample_ppc(trace, samples=500, model=model, size=100)
+                if len(ppc) == 0:
+                    npt.assert_true(len(ppc) == 0, 'length of ppc sample is zero')
+            except:
+                assert False
+