Merge pull request #1605 from lamblin/scan_minor

theano/sandbox/scan.py

@@ -196,9 +196,13 @@ def scan(fn,
     inner_slices = []  # Actual slices if scan is removed from the picture
     # go through sequences picking up time slices as needed
     for i, seq in enumerate(seqs):
+        if isinstance(seq, dict):
+            seq = seq['input']
         actual_slice = seq[0]
         _seq_val = tensor.as_tensor_variable(seq)
         _seq_val_slice = _seq_val[0]
+
+        nw_slice = _seq_val_slice.type()
         # Try to transfer test_value to the new variable
         if config.compute_test_value != 'off':
             try:
@@ -212,7 +216,6 @@ def scan(fn,
                         'the inner function of scan, input value '
                         'missing %s'), e)
 
-        nw_slice = _seq_val_slice.type()
         if seq.name:
             nw_slice.name = seq.name + '[t]'
         scan_seqs.append(_seq_val)

theano/scan_module/scan.py

@@ -384,7 +384,7 @@ def scan(fn,
             seqs[i] = OrderedDict([('input', seqs[i]), ('taps', [0])])
         elif seqs[i].get('taps', None):
             seqs[i]['taps'] = wrap_into_list(seqs[i]['taps'])
-        elif seqs[i].get('taps', True) is None:
+        elif seqs[i].get('taps', None) is None:
             # seqs dictionary does not have the ``taps`` key
             seqs[i]['taps'] = [0]
 

theano/scan_module/scan_opt.py

@@ -1451,51 +1451,15 @@ def scan_merge_inouts(node):
         seen.append((outer_i, inner_o, outer_o))
         return outer_o
 
-    def map_nitsot_out(outer_i, inner_o, outer_o, sh, seen):
-        # Like map_out, but also checks the needed shape.
-        for p, (s_outer_i, s_inner_o, s_outer_o, ssh) in enumerate(seen):
-            if (equal_computations([inner_o], [s_inner_o], left, right)
-                    and outer_i == s_outer_i):
-                if equal_computations([sh], [ssh]):
-                    return s_outer_o
-                try:
-                    vsh = int(opt.get_scalar_constant_value(sh))
-                    vssh = int(opt.get_scalar_constant_value(ssh))
-                except tensor.NotScalarConstantError:
-                    return outer_o
-                if vsh == vssh:
-                    return s_outer_o
-                elif vsh > vssh:
-                    seen[p] = (outer_i, inner_o, outer_o, sh)
-                    return outer_o
-                else:
-                    return s_outer_o[:vsh]
-        seen.append((outer_i, inner_o, outer_o, sh))
-        return outer_o
-
     seen = []
 
-    shapes = []
-    for x in na.outer_in_nit_sot:
-        if x.ndim > 0:
-            if hasattr(node.fgraph, 'shape_feature'):
-                shapes.append(
-                    node.fgraph.shape_feature.shape_of[x][0])
-            else:
-                shapes.append(x.shape[0])
-        else:
-            # If x is a scalar, then it means its value is the number of
-            # items scan is supposed to store for this nit_sot sequence
-            shapes.append(x)
     assert len(na.outer_in_nit_sot) == len(na.inner_out_nit_sot)
     assert len(na.inner_out_nit_sot) == len(na.outer_out_nit_sot)
-    assert len(na.outer_out_nit_sot) == len(shapes)
     na.outer_out_nit_sot = [
-        map_nitsot_out(outer_i, inner_o, outer_o, sh, seen)
-        for outer_i, inner_o, outer_o, sh in zip(na.outer_in_nit_sot,
-                                                 na.inner_out_nit_sot,
-                                                 na.outer_out_nit_sot,
-                                                 shapes)]
+        map_out(outer_i, inner_o, outer_o, seen)
+        for outer_i, inner_o, outer_o in zip(na.outer_in_nit_sot,
+                                             na.inner_out_nit_sot,
+                                             na.outer_out_nit_sot)]
 
     seen = []
     assert len(na.outer_in_sit_sot) == len(na.inner_out_sit_sot)

theano/scan_module/tests/test_scan.py

@@ -2477,6 +2477,18 @@ class T_Scan(unittest.TestCase):
         # Run it so DebugMode can detect optimization problems.
         f(x_val, y_val)
 
+    def test_sequence_dict(self):
+        # Test that we can specify sequences as a dictionary with
+        # only the 'input' key
+        def incr(s):
+            return s + 1
+
+        x = theano.tensor.vector()
+        sx, upx = theano.scan(
+            fn=incr,
+            sequences=[{'input': x}])
+        f = theano.function([x], sx)
+
     def test_hash(self):
         x = theano.tensor.vector()
         y = theano.tensor.vector()