change the default of include_transformed... (#2379)

to `True`. Users confusion in sample_ppc or sample_gp could be avoide (e.g., Users are reporting error: #2368 and https://discourse.pymc.io/t/bayesian-neural-net-with-beta-distribution/136)

Author: Junpeng Lao

pymc3/variational/opvi.py

@@ -878,7 +878,7 @@ def inner(draws=1):
             return dict([(n_, s_) for n_, s_ in zip(self.global_names+self.local_names, _samples)])
         return inner
 
-    def sample(self, draws=500, include_transformed=False):
+    def sample(self, draws=500, include_transformed=True):
         """Draw samples from variational posterior.
 
         Parameters