Merge pull request #124 from bartvm/more_fixes

Make Cost and CostMatrix abstract classses

blocks/bricks/cost.py

@@ -1,16 +1,23 @@
+from abc import ABCMeta, abstractmethod
+
 import theano
 from theano import tensor
+from six import add_metaclass
 
 from blocks.bricks import application, Brick
 
-
 floatX = theano.config.floatX
 
 
+@add_metaclass(ABCMeta)
 class Cost(Brick):
-    pass
+    @abstractmethod
+    @application
+    def apply(self, y, y_hat):
+        pass
 
 
+@add_metaclass(ABCMeta)
 class CostMatrix(Cost):
     """Base class for costs which can be calculated element-wise.
 
@@ -22,6 +29,11 @@ def apply(self, y, y_hat):
         return self.cost_matrix.application_method(
             self, y, y_hat).sum(axis=1).mean()
 
+    @abstractmethod
+    @application
+    def cost_matrix(self, y, y_hat):
+        pass
+
 
 class BinaryCrossEntropy(CostMatrix):
     @application

blocks/bricks/recurrent.py

@@ -463,7 +463,7 @@ def __init__(self, prototype, **kwargs):
         super(Bidirectional, self).__init__(**kwargs)
         self.prototype = prototype
 
-        self.children = [copy.deepcopy(prototype) for i in range(2)]
+        self.children = [copy.deepcopy(prototype) for _ in range(2)]
         self.children[0].name = 'forward'
         self.children[1].name = 'backward'
 

blocks/bricks/sequence_generators.py

@@ -509,7 +509,7 @@ def _push_allocation_config(self):
         self.lookup.dim = self.feedback_dim
 
     @application
-    def feedback(self, outputs, **kwargs):
+    def feedback(self, outputs):
         assert self.output_dim == 0
         return self.lookup.lookup(outputs)
 

blocks/datasets/__init__.py

@@ -39,6 +39,8 @@ class Dataset(object):
     simultaneously.
 
     """
+    sources = None
+
     def __init__(self, sources=None):
         if sources is not None:
             if not all(source in self.sources for source in sources):

blocks/datasets/mnist.py

@@ -153,7 +153,7 @@ def read_mnist_labels(filename):
 
     """
     with open(filename, 'rb') as f:
-        magic, number = struct.unpack('>ii', f.read(8))
+        magic, _ = struct.unpack('>ii', f.read(8))
         if magic != MNIST_LABEL_MAGIC:
             raise ValueError("Wrong magic number reading MNIST label file")
         array = numpy.fromfile(f, dtype='uint8')

blocks/groundhog.py

@@ -74,11 +74,11 @@ def inputs(self):
 
     @property
     def valid_costs(self):
-        return ['cost'] + [name for name, var in self.properties]
+        return ['cost'] + [name for name, _ in self.properties]
 
     def validate(self, data):
         valid_names = self.valid_costs
-        valid_vars = [self.train_cost] + [var for name, var in self.properties]
+        valid_vars = [self.train_cost] + [var for _, var in self.properties]
 
         sums = numpy.zeros((len(valid_vars),))
         num_batches = 0
@@ -115,7 +115,7 @@ def load(self, path):
 
 class GroundhogState(object):
     """Good default values for groundhog state."""
-    def __init__(self, prefix, batch_size, learning_rate, **kwargs):
+    def __init__(self, prefix, batch_size, learning_rate):
         self.prefix = prefix
         self.bs = batch_size
         self.lr = learning_rate

blocks/utils.py

@@ -360,7 +360,7 @@ def repr_attrs(instance, *attrs):
     orig_repr_template += '>'
     try:
         return repr_template.format(instance, id(instance))
-    except:
+    except Exception:
         return orig_repr_template.format(instance, id(instance))