Merge pull request #479 from nouiz/except

theano/gof/env.py

@@ -443,7 +443,7 @@ class Env(utils.object2):
         """
         try:
             self._features.remove(feature)
-        except:
+        except Exception:
             return
         detach = getattr(feature, 'on_detach', None)
         if detach is not None:

theano/gof/tests/test_optdb.py

-from theano.gof.optdb import *
 from unittest import TestCase
 
+from theano.gof.optdb import opt, DB
+
+
 class Test_DB(TestCase):
 
     def test_0(self):
 
-        class Opt(opt.Optimizer):  #inheritance buys __hash__
+        class Opt(opt.Optimizer):  # inheritance buys __hash__
             name = 'blah'
 
         db = DB()
@@ -16,36 +18,34 @@ class Test_DB(TestCase):
         db.register('c', Opt(), 'z', 'asdf')
 
         try:
-            db.register('c', Opt()) #name taken
+            db.register('c', Opt())  # name taken
             self.fail()
         except ValueError, e:
             if e[0].startswith("The name"):
                 pass
             else:
                 raise
-        except:
+        except Exception:
             self.fail()
 
         try:
-            db.register('z', Opt()) #name collides with tag
+            db.register('z', Opt())  # name collides with tag
             self.fail()
         except ValueError, e:
             if e[0].startswith("The name"):
                 pass
             else:
                 raise
-        except:
+        except Exception:
             self.fail()
 
         try:
-            db.register('u', Opt(), 'b') #name new but tag collides with name
+            db.register('u', Opt(), 'b')  # name new but tag collides with name
             self.fail()
         except ValueError, e:
             if e[0].startswith("The tag"):
                 pass
             else:
                 raise
-        except:
+        except Exception:
             self.fail()
-
-

theano/gof/utils.py

@@ -285,20 +285,11 @@ def comm_guard(type1, type2):
                      and (type2 is ANY_TYPE or isinstance(arg1, type2)):
                 arg1, arg2 = arg2, arg1
             else:
-                try:
-                    return old_f(arg1, arg2, *rest)
-                except:
-                    raise
-
-            try:
-                variable = f(arg1, arg2, *rest)
-            except:
-                raise
+                return old_f(arg1, arg2, *rest)
+
+            variable = f(arg1, arg2, *rest)
             if variable is FALL_THROUGH:
-                try:
-                    return old_f(arg1, arg2, *rest)
-                except:
-                    raise
+                return old_f(arg1, arg2, *rest)
             else:
                 return variable
 

theano/sandbox/scan_module/tests/test_scan.py

@@ -299,13 +299,13 @@ class TestScan(unittest.TestCase):
                 for th_out, num_out in zip(theano_outs, numpy_outs):
                     try:
                         assert numpy.allclose(th_out, num_out)
-                    except:
+                    except Exception:
                         #import ipdb; ipdb.set_trace()
                         raise
                 for th_out, num_out in zip(shared_vars, numpy_shared):
                     try:
                         assert numpy.allclose(th_out.get_value(), num_out)
-                    except:
+                    except Exception:
                         #import ipdb; ipdb.set_trace()
                         raise
                 # Scenario 2 : Loose fit (sequences longer then required)

theano/tensor/nnet/tests/test_conv.py

@@ -310,25 +310,12 @@ class TestConv2D(unittest.TestCase):
         """
         Make sure errors are raised when image and kernel are not 4D tensors
         """
-        try:
-            self.validate((3,2,8,8), (4,2,5,5), 'valid', input = T.dmatrix())
-            # should never reach here
-            self.fail()
-        except:
-            pass
-        try:
-            self.validate((3,2,8,8), (4,2,5,5), 'valid', filters = T.dvector())
-            # should never reach here
-            self.fail()
-        except:
-            pass
-        try:
-            self.validate((3,2,8,8), (4,2,5,5), 'valid', input = T.dtensor3())
-            # should never reach here
-            self.fail()
-        except:
-            pass
-
+        self.assertRaises(Exception, self.validate, (3, 2, 8, 8), (4, 2, 5, 5),
+                          'valid', input=T.dmatrix())
+        self.assertRaises(Exception, self.validate, (3, 2, 8, 8), (4, 2, 5, 5),
+                          'valid', filters=T.dvector())
+        self.assertRaises(Exception, self.validate, (3, 2, 8, 8), (4, 2, 5, 5),
+                          'valid', input=T.dtensor3())
 
     def test_gcc_crash(self):
         """

theano/tensor/nnet/tests/test_nnet.py

@@ -496,7 +496,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
             try:
                 assert len(f.maker.env.toposort()) == 4
                 f(x_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(f)
                 raise
 
@@ -507,7 +507,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
             try:
                 assert len(g.maker.env.toposort()) == 4
                 g(x_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(g)
                 raise
 
@@ -525,7 +525,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
             try:
                 assert len(f.maker.env.toposort()) == 2 # [big_op, sum]
                 f(x_val, b_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(f)
                 raise
 
@@ -535,7 +535,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
             try:
                 assert len(g.maker.env.toposort()) == 4
                 g(x_val, b_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(g)
                 raise
 
@@ -553,7 +553,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
             try:
                 assert len(f.maker.env.toposort()) == 6
                 f(x_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(f)
                 raise
 
@@ -564,7 +564,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
                 assert len(g.maker.env.toposort()) in (6,7) #there's an extra dimshuffle in there
                 # but I can't think of a good rule to get rid of it
                 g(x_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(g)
                 raise
 
@@ -580,7 +580,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
                 theano.printing.debugprint(f)
             try:
                 assert len(f.maker.env.toposort()) == 4
-            except:
+            except Exception:
                 theano.printing.debugprint(f)
                 raise
 
@@ -590,7 +590,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
             try:
                 assert len(g.maker.env.toposort()) in (6,7)
                 g(x_val, b_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(g)
                 raise
 
@@ -624,7 +624,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
             try:
                 assert len(f.maker.env.toposort()) == 5
                 f(x_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(f)
                 raise
 
@@ -635,7 +635,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
             try:
                 assert len(g.maker.env.toposort()) == 5
                 g(x_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(g)
                 raise
 
@@ -671,7 +671,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
                 prev, last = f.maker.env.toposort()[-2:]
                 assert len(f.maker.env.toposort()) == 5
                 f(x_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(f)
                 raise
 
@@ -684,7 +684,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
                 assert softmax in ops
                 assert softmax_grad not in ops
                 g(x_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(g)
                 raise
 
@@ -723,7 +723,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
                 prev, last = f.maker.env.toposort()[-2:]
                 assert len(f.maker.env.toposort()) == 3 # [big_op, sum, dim_shuffle]
                 f(x_val, b_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(f)
                 raise
 
@@ -742,7 +742,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
                 assert softmax_with_bias in ops
                 assert softmax_grad not in ops
                 g(x_val, b_val, y_val)
-            except:
+            except Exception:
                 theano.printing.debugprint(g)
                 raise
 
@@ -831,7 +831,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
                 assert 5 <= len(f.maker.env.toposort()) <= 10
                 validate_fn_graph(f)
                 f(x_val, y_val, 0.1)
-            except:
+            except Exception:
                 theano.printing.debugprint(f)
                 raise
 
@@ -841,7 +841,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
                 assert 5 <= len(g.maker.env.toposort()) <= 12
                 validate_grad_graph(g)
                 g(x_val, y_val, 0.1)
-            except:
+            except Exception:
                 theano.printing.debugprint(g)
                 raise
 
@@ -851,7 +851,7 @@ class T_CrossentropyCategorical1Hot(unittest.TestCase):
                 assert 8 <= len(h.maker.env.toposort()) <= 17
                 validate_grad_graph(h)
                 h(x_val, y_val, 0.1)
-            except:
+            except Exception:
                 theano.printing.debugprint(h)
                 raise
 

theano/tensor/sharedvar.py

 import traceback
+
 import numpy
+
 import theano.tensor.basic
-from basic import TensorType, _tensor_py_operators, autocast_int, autocast_float
+from basic import TensorType, _tensor_py_operators
 from theano.compile import shared_constructor, SharedVariable
-from theano import config
+
 
 def load_shared_variable(val):
     """This function is only here to keep some pickles loading
@@ -11,35 +13,40 @@ def load_shared_variable(val):
     It can be removed after sufficient time has passed."""
     return tensor_constructor(val)
 
+
 # _tensor_py_operators is first to have its version of __{gt,ge,lt,le}__
 class TensorSharedVariable(_tensor_py_operators, SharedVariable):
     pass
 
+
 @shared_constructor
-def tensor_constructor(value, name=None, strict=False, allow_downcast=None, borrow=False, broadcastable=None):
+def tensor_constructor(value, name=None, strict=False, allow_downcast=None,
+                       borrow=False, broadcastable=None):
     """SharedVariable Constructor for TensorType
 
     :note: Regarding the inference of the broadcastable pattern...
-    The default is to assume that the value might be resized in any dimension, so the default
-    broadcastable is ``(False,)*len(value.shape)``.  The optional `broadcastable` argument will
-    override this default.
+    The default is to assume that the value might be resized in any
+    dimension, so the default broadcastable is
+    ``(False,)*len(value.shape)``.  The optional `broadcastable`
+    argument will override this default.
 
     """
     if not isinstance(value, numpy.ndarray):
         raise TypeError()
 
-    # if no broadcastable is given, then the default is to assume that the value might be
-    # resized in any dimension in the future.
+    # if no broadcastable is given, then the default is to assume that
+    # the value might be resized in any dimension in the future.
     #
     if broadcastable is None:
-        broadcastable = (False,)*len(value.shape)
+        broadcastable = (False,) * len(value.shape)
     type = TensorType(value.dtype, broadcastable=broadcastable)
     return TensorSharedVariable(type=type,
-            value=numpy.array(value,copy=(not borrow)),
+            value=numpy.array(value, copy=(not borrow)),
             name=name,
             strict=strict,
             allow_downcast=allow_downcast)
 
+
 # TensorSharedVariable brings in the tensor operators, is not ideal, but works
 # as long as we dont do purely scalar-scalar operations
 # _tensor_py_operators is first to have its version of __{gt,ge,lt,le}__
@@ -50,6 +57,7 @@ def tensor_constructor(value, name=None, strict=False, allow_downcast=None, borr
 class ScalarSharedVariable(_tensor_py_operators, SharedVariable):
     pass
 
+
 @shared_constructor
 def scalar_constructor(value, name=None, strict=False, allow_downcast=None):
     """SharedVariable constructor for scalar values. Default: int64 or float64.
@@ -57,14 +65,14 @@ def scalar_constructor(value, name=None, strict=False, allow_downcast=None):
     :note: We implement this using 0-d tensors for now.
 
     """
-    if not isinstance (value, (numpy.number, float, int, complex)):
+    if not isinstance(value, (numpy.number, float, int, complex)):
         raise TypeError()
     try:
-        dtype=value.dtype
-    except:
-        dtype=numpy.asarray(value).dtype
+        dtype = value.dtype
+    except Exception:
+        dtype = numpy.asarray(value).dtype
 
-    dtype=str(dtype)
+    dtype = str(dtype)
     value = theano._asarray(value, dtype=dtype)
     tensor_type = TensorType(dtype=str(value.dtype), broadcastable=[])
 
@@ -75,6 +83,6 @@ def scalar_constructor(value, name=None, strict=False, allow_downcast=None):
                 value=numpy.array(value, copy=True),
                 name=name, strict=strict, allow_downcast=allow_downcast)
         return rval
-    except:
+    except Exception:
         traceback.print_exc()
         raise

theano/tensor/signal/tests/test_conv.py

-import sys, time, unittest
+import unittest
+
 import numpy
 
 import theano
 import theano.tensor as T
-from theano import function, Mode
 from theano.tests import unittest_tools as utt
 
 from theano.tensor.signal import conv
 
 from theano.tensor.basic import _allclose
 
+
 class TestSignalConv2D(unittest.TestCase):
 
     def setUp(self):
@@ -19,13 +20,15 @@ class TestSignalConv2D(unittest.TestCase):
 
         image_dim = len(image_shape)
         filter_dim = len(filter_shape)
-        input = T.TensorType('float64', [False]*image_dim)()
-        filters = T.TensorType('float64', [False]*filter_dim)()
+        input = T.TensorType('float64', [False] * image_dim)()
+        filters = T.TensorType('float64', [False] * filter_dim)()
 
         bsize = image_shape[0]
-        if image_dim!=3: bsize = 1
+        if image_dim != 3:
+            bsize = 1
         nkern = filter_shape[0]
-        if filter_dim!=3: nkern = 1
+        if filter_dim != 3:
+            nkern = 1
 
         ############# THEANO IMPLEMENTATION ############
         # we create a symbolic function so that verify_grad can work
@@ -35,7 +38,7 @@ class TestSignalConv2D(unittest.TestCase):
         theano_conv = theano.function([input, filters], output)
 
         # initialize input and compute result
-        image_data  = numpy.random.random(image_shape)
+        image_data = numpy.random.random(image_shape)
         filter_data = numpy.random.random(filter_shape)
         theano_output = theano_conv(image_data, filter_data)
 
@@ -45,10 +48,11 @@ class TestSignalConv2D(unittest.TestCase):
         ref_output = numpy.zeros(tuple(out_shape2d))
 
         # reshape as 3D input tensors to make life easier
-        image_data3d  = image_data.reshape((bsize,)+image_shape[-2:])
-        filter_data3d = filter_data.reshape((nkern,)+filter_shape[-2:])
+        image_data3d = image_data.reshape((bsize,) + image_shape[-2:])
+        filter_data3d = filter_data.reshape((nkern,) + filter_shape[-2:])
         # reshape theano output as 4D to make life easier
-        theano_output4d = theano_output.reshape((bsize,nkern,)+theano_output.shape[-2:])
+        theano_output4d = theano_output.reshape((bsize, nkern,) +
+                                                theano_output.shape[-2:])
 
         # loop over mini-batches (if required)
         for b in range(bsize):
@@ -56,17 +60,19 @@ class TestSignalConv2D(unittest.TestCase):
             # loop over filters (if required)
             for k in range(nkern):
 
-                image2d = image_data3d[b,:,:]
-                filter2d = filter_data3d[k,:,:]
+                image2d = image_data3d[b, :, :]
+                filter2d = filter_data3d[k, :, :]
                 output2d = numpy.zeros(ref_output.shape)
                 for row in range(ref_output.shape[0]):
                     for col in range(ref_output.shape[1]):
-                        output2d[row,col] += (image2d[row:row+filter2d.shape[0],
-                                                            col:col+filter2d.shape[1]]*filter2d[::-1,::-1]
-                                                    ).sum()
+                        output2d[row, col] += (
+                            image2d[row:row + filter2d.shape[0],
+                                    col:col + filter2d.shape[1]] *
+                            filter2d[::-1, ::-1]
+                            ).sum()
 
-
-                self.assertTrue(_allclose(theano_output4d[b,k,:,:], output2d))
+                self.assertTrue(_allclose(theano_output4d[b, k, :, :],
+                                          output2d))
 
         ############# TEST GRADIENT ############
         if verify_grad:
@@ -74,29 +80,22 @@ class TestSignalConv2D(unittest.TestCase):
 
     def test_basic(self):
         """
-        Basic functionality of nnet.conv.ConvOp is already tested by its own test suite.  We
-        just have to test whether or not signal.conv.conv2d can support inputs and filters of
-        type matrix or tensor3.
+        Basic functionality of nnet.conv.ConvOp is already tested by
+        its own test suite.  We just have to test whether or not
+        signal.conv.conv2d can support inputs and filters of type
+        matrix or tensor3.
         """
-        self.validate((1,4,5), (2,2,3), verify_grad=True)
-        self.validate((7,5), (5,2,3), verify_grad=False)
-        self.validate((3,7,5), (2,3), verify_grad=False)
-        self.validate((7,5), (2,3), verify_grad=False)
+        self.validate((1, 4, 5), (2, 2, 3), verify_grad=True)
+        self.validate((7, 5), (5, 2, 3), verify_grad=False)
+        self.validate((3, 7, 5), (2, 3), verify_grad=False)
+        self.validate((7, 5), (2, 3), verify_grad=False)
 
     def test_fail(self):
         """
         Test that conv2d fails for dimensions other than 2 or 3.
         """
-        try:
-            conv.conv2d(T.dtensor4(), T.dtensor3())
-            self.fail()
-        except:
-            pass
-        try:
-            conv.conv2d(T.dtensor3(), T.dvector())
-            self.fail()
-        except:
-            pass
+        self.assertRaises(Exception, conv.conv2d, T.dtensor4(), T.dtensor3())
+        self.assertRaises(Exception, conv.conv2d, T.dtensor3(), T.dvector())
 
     def test_bug_josh_reported(self):
         """
@@ -106,5 +105,4 @@ class TestSignalConv2D(unittest.TestCase):
         """
         m1 = theano.tensor.matrix()
         m2 = theano.tensor.matrix()
-        rval = conv.conv2d(m1, m2)
-
+        conv.conv2d(m1, m2)