pyro/infer/predictive.py

Survived

Mutant 338

--- pyro/infer/predictive.py
+++ pyro/infer/predictive.py
@@ -31,7 +31,7 @@
 
 
 def _predictive_sequential(model, posterior_samples, model_args, model_kwargs,
-                           num_samples, return_site_shapes, return_trace=False):
+                           num_samples, return_site_shapes, return_trace=True):
     collected = []
     samples = [{k: v[i] for k, v in posterior_samples.items()} for i in range(num_samples)]
     for i in range(num_samples):

Mutant 339

--- pyro/infer/predictive.py
+++ pyro/infer/predictive.py
@@ -49,7 +49,7 @@
 
 
 def _predictive(model, posterior_samples, num_samples, return_sites=(),
-                return_trace=False, parallel=False, model_args=(), model_kwargs={}):
+                return_trace=True, parallel=False, model_args=(), model_kwargs={}):
     max_plate_nesting = _guess_max_plate_nesting(model, model_args, model_kwargs)
     vectorize = pyro.plate("_num_predictive_samples", num_samples, dim=-max_plate_nesting-1)
     model_trace = prune_subsample_sites(poutine.trace(model).get_trace(*model_args, **model_kwargs))

Mutant 340

--- pyro/infer/predictive.py
+++ pyro/infer/predictive.py
@@ -49,7 +49,7 @@
 
 
 def _predictive(model, posterior_samples, num_samples, return_sites=(),
-                return_trace=False, parallel=False, model_args=(), model_kwargs={}):
+                return_trace=False, parallel=True, model_args=(), model_kwargs={}):
     max_plate_nesting = _guess_max_plate_nesting(model, model_args, model_kwargs)
     vectorize = pyro.plate("_num_predictive_samples", num_samples, dim=-max_plate_nesting-1)
     model_trace = prune_subsample_sites(poutine.trace(model).get_trace(*model_args, **model_kwargs))

Mutant 341

--- pyro/infer/predictive.py
+++ pyro/infer/predictive.py
@@ -133,7 +133,7 @@
         all batch dims correctly annotated via :class:`~pyro.plate`. Default is `False`.
     """
     def __init__(self, model, posterior_samples=None, guide=None, num_samples=None,
-                 return_sites=(), parallel=False):
+                 return_sites=(), parallel=True):
         super().__init__()
         if posterior_samples is None:
             if num_samples is None: