Alter PdbBreakpoint to show the user modifiables cpu-versions of the monitored variables

theano/tests/breakpoint.py

+import numpy
 import pdb
 
 import theano
@@ -78,14 +79,17 @@ class PdbBreakpoint(Op):
 
         # Because the user might be tempted to instantiate PdbBreakpoint only
         # once and apply it many times on different number of inputs, we must
-        # create a new instance of the op here, define the view_map in that
-        # instance and then apply it on the inputs.
+        # create a new instance of the op here, define the instance attributes
+        # (view_map and var_types) in that instance and then apply it on the
+        # inputs.
         new_op = PdbBreakpoint(name=self.name)
         new_op.view_map = {}
+        new_op.inp_types = []
         for i in range(len(monitored_vars)):
             # Every output i is a view of the input i+1 because of the input
             # condition.
             new_op.view_map[i] = [i+1]
+            new_op.inp_types.append(monitored_vars[i].type)
 
         # Build the Apply node
         inputs = [condition] + list(monitored_vars)
@@ -94,18 +98,28 @@ class PdbBreakpoint(Op):
 
     def perform(self, node, inputs, output_storage):
         condition = inputs[0]
-        monitored = inputs[1:]
+        try:
+            monitored = [numpy.asarray(inp) for inp in inputs[1:]]
+        except:
+            raise ValueError("Some of the inputs to the PdbBreakpoint op '%s'"
+                             "could not be casted to NumPy arrays" %
+                             self.name)
 
         if condition:
-            print "-------------------------------------------------"
-            print "Conditional breakpoint %s activated" % self.name
-            print "The monitored variables are stored, in order,"
-            print "in the list variable 'monitored'"
-            print "-------------------------------------------------"
+            print("\n")
+            print("-------------------------------------------------")
+            print("Conditional breakpoint '%s' activated\n" % self.name)
+            print("The monitored variables are stored, in order,")
+            print("in the list variable 'monitored' as NumPy arrays.\n")
+            print("Their contents can be altered and, when execution")
+            print("resumes, the updated values will be used.")
+            print("-------------------------------------------------")
             pdb.set_trace()
 
+        # Take the new values in monitored, cast them back to their original
+        # type and store them in the output_storage
         for i in range(len(output_storage)):
-            output_storage[i][0] = monitored[i]
+            output_storage[i][0] = self.inp_types[i].filter(monitored[i])
 
     def grad(self, inputs, output_gradients):
         return ([DisconnectedType()()] + output_gradients)