Merge pull request #4378 from shabanian/bn

theano/gof/op.py

@@ -13,7 +13,6 @@ import numpy
 import os
 import re
 import sys
-import traceback
 import warnings
 
 import theano
@@ -21,7 +20,6 @@ from theano import config
 
 import theano.gof.cc
 from six import itervalues
-from six.moves import StringIO
 from theano.gof import graph
 from theano.gof import utils
 from theano.gof.cmodule import GCC_compiler
@@ -554,16 +552,7 @@ class PureOp(object):
                 detailed_err_msg = (
                     "For compute_test_value, one input test value does not"
                     " have the requested type.\n")
-                tr = getattr(v.tag, 'trace', [])
-                if isinstance(tr, list) and len(tr) > 0:
-                    detailed_err_msg += (
-                        " \nBacktrace when that variable is created:\n")
-                    # Print separate message for each element in the list
-                    # of batcktraces
-                    sio = StringIO()
-                    for subtr in tr:
-                        traceback.print_list(subtr, sio)
-                    detailed_err_msg += str(sio.getvalue())
+                detailed_err_msg += utils.get_variable_trace_string(v)
 
                 detailed_err_msg += (
                     "\nThe error when converting the test value to that"
@@ -575,8 +564,8 @@ class PureOp(object):
                 e.args = ("\n".join(args),)
                 raise
             return ret
-
-        raise AttributeError('%s has no test value' % v)
+        detailed_err_msg = utils.get_variable_trace_string(v)
+        raise AttributeError('%s has no test value %s' % (v, detailed_err_msg))
 
     def __call__(self, *inputs, **kwargs):
         """
@@ -630,9 +619,11 @@ class PureOp(object):
                             (i, ins, node), stacklevel=2)
                         run_perform = False
                     elif config.compute_test_value == 'raise':
+                        detailed_err_msg = utils.get_variable_trace_string(ins)
+
                         raise ValueError(
-                            'Cannot compute test value: input %i (%s) of Op %s missing default value' %
-                            (i, ins, node))
+                            'Cannot compute test value: input %i (%s) of Op %s missing default value. %s' %
+                            (i, ins, node, detailed_err_msg))
                     elif config.compute_test_value == 'ignore':
                         # silently skip test
                         run_perform = False

theano/gof/utils.py

 from __future__ import absolute_import, print_function, division
 import linecache
 import sys
+import traceback
 
 import numpy
 from six import iteritems, integer_types, string_types
+from six.moves import StringIO
 
 from theano import config
 from theano.compat import OrderedDict, PY3
@@ -99,6 +101,7 @@ def add_tag_trace(thing, user_line=None):
              "theano/scan_module/", "theano\\scan_module\\",
              "theano/sparse/", "theano\\sparse\\",
              "theano/typed_list/", "theano\\typed_list\\"]
+
     tr = simple_extract_stack(limit=user_line, skips=skips)
     # Different python version use different sementic for
     # limit. python 2.7 include the call to extrack_stack. The -1 get
@@ -111,6 +114,23 @@ def add_tag_trace(thing, user_line=None):
     return thing
 
 
+def get_variable_trace_string(v):
+    sio = StringIO()
+    # For backward compatibility with old trace
+    tr = getattr(v.tag, 'trace', [])
+    if isinstance(tr, list) and len(tr) > 0:
+        print(" \nBacktrace when that variable is created:\n", file=sio)
+        # The isinstance is needed to handle old pickled trace
+        if isinstance(tr[0], tuple):
+            traceback.print_list(v.tag.trace, sio)
+        else:
+            # Print separate message for each element in the list of
+            # batcktraces
+            for subtr in tr:
+                traceback.print_list(subtr, sio)
+    return sio.getvalue()
+
+
 def hashtype(self):
     t = type(self)
     return hash(t.__name__) ^ hash(t.__module__)

theano/gradient.py

@@ -3,17 +3,15 @@ from __future__ import absolute_import, print_function, division
 import six.moves.builtins as builtins
 import logging
 import time
-import traceback
 import warnings
 
 import numpy  # for numeric_grad
 from six import itervalues
-from six.moves import StringIO
 
 import theano
 
 from theano import gof
-from theano.gof import Variable
+from theano.gof import utils, Variable
 from theano.compat import OrderedDict, izip
 from six.moves import xrange, reduce
 from theano.gof.null_type import NullType, null_type
@@ -518,17 +516,7 @@ def grad(cost, wrt, consider_constant=None,
             elif disconnected_inputs == 'warn':
                 warnings.warn(message, stacklevel=2)
             elif disconnected_inputs == 'raise':
-                # Add the var trace
-                tr = getattr(var.tag, 'trace', [])
-                if len(tr) > 0:
-                    message += "\nBacktrace when the node is created:\n"
-
-                    # Print separate message for each element in the list of batcktraces
-                    sio = StringIO()
-                    for subtr in tr:
-                        traceback.print_list(subtr, sio)
-                    message += str(sio.getvalue())
-
+                message = utils.get_variable_trace_string(var)
                 raise DisconnectedInputError(message)
             else:
                 raise ValueError("Invalid value for keyword "

theano/sandbox/rng_mrg.py

@@ -1205,6 +1205,7 @@ class MRG_RandomStreams(object):
         self.rstate = multMatVect(self.rstate, A1p134, M1, A2p134, M2)
         assert self.rstate.dtype == numpy.int32
 
+    @theano.configparser.change_flags(compute_test_value='off')
     def get_substream_rstates(self, n_streams, dtype, inc_rstate=True):
         # TODO : need description for parameter and return
         """

theano/sandbox/tests/test_rng_mrg.py

@@ -109,6 +109,21 @@ def test_consistency_randomstreams():
         assert(numpy.allclose(samples, java_samples))
 
 
+def test_get_substream_rstates():
+    try:
+        orig = theano.config.compute_test_value
+        theano.config.compute_test_value = 'raise'
+        n_streams = 100
+
+        dtype = 'float32'
+        rng = MRG_RandomStreams(numpy.random.randint(2147462579))
+
+        rng.get_substream_rstates(n_streams, dtype)
+
+    finally:
+        theano.config.compute_test_value = orig
+
+
 def test_consistency_cpu_serial():
     """
     Verify that the random numbers generated by mrg_uniform, serially,

theano/tensor/nnet/bn.py

@@ -7,6 +7,7 @@ from theano.scalar import add, sub, true_div, mul
 class BNComposite(Composite):
     init_param = ('dtype',)
 
+    @theano.configparser.change_flags(compute_test_value='off')
     def __init__(self, dtype):
         self.dtype = dtype
         x = theano.scalar.Scalar(dtype=dtype).make_variable()

theano/tensor/nnet/tests/test_bn.py

@@ -6,6 +6,47 @@ import numpy
 from theano.tensor.nnet.bn import batch_normalization
 
 
+def test_BNComposite():
+    try:
+        orig = theano.config.compute_test_value
+
+        theano.config.compute_test_value = 'raise'
+
+        def bn_ref(x, G, B, M, V):
+            n = (x - M) / V
+            return n * G + B
+
+        numpy.random.seed(1234)
+        X = 1 + numpy.random.random([10, 20]).astype('float32')
+        B = 1 + numpy.random.random([20]).astype('float32')
+        G = 1 + numpy.random.random([20]).astype('float32')
+        M = 1 + numpy.random.random([20]).astype('float32')
+        V = 1 + numpy.random.random([20]).astype('float32')
+
+        x = theano.tensor.matrix('x')
+        b = theano.tensor.vector('b')
+        g = theano.tensor.vector('g')
+        m = theano.tensor.vector('m')
+        v = theano.tensor.vector('v')
+
+        x.tag.test_value = numpy.random.rand(2, 2).astype(theano.config.floatX)
+        b.tag.test_value = numpy.random.rand(2).astype(theano.config.floatX)
+        g.tag.test_value = numpy.random.rand(2).astype(theano.config.floatX)
+        m.tag.test_value = numpy.random.rand(2).astype(theano.config.floatX)
+        v.tag.test_value = numpy.random.rand(2).astype(theano.config.floatX)
+
+        bn_ref_op = bn_ref(x, g, b, m, v)
+        f_ref = theano.function([x, b, g, m, v], [bn_ref_op])
+        res_ref = f_ref(X, G, B, M, V)
+        for mode in ['low_mem', 'high_mem']:
+            bn_op = batch_normalization(x, g, b, m, v, mode=mode)
+            f = theano.function([x, b, g, m, v], [bn_op])
+            res = f(X, G, B, M, V)
+            utt.assert_allclose(res_ref, res)
+    finally:
+        theano.config.compute_test_value = orig
+
+
 def test_bn():
 
     def bn_ref(x, G, B, M, V):