https://github.com/pandas-dev/pandas/issues/15091
I fixed it by passing a numpy array. Ugly, but it works.

c_2_o = pd.DataFrame()
h_2_o = pd.DataFrame()
l_2_o = pd.DataFrame()
c_2_h = pd.DataFrame()
h_2_l = pd.DataFrame()
c1_c0 = pd.DataFrame()
vol = pd.DataFrame()

def make_inputs(filepath):
    D = pd.read_csv(filepath,header=None,names=['ticker','o','h','l','c','v']) 
    #Load the dataframe with headers
    D.index = pd.to_datetime(D.index,format='%Y%m%d') 
    # Set the indix to a datetime
    ticker = str(get_ticker(filepath))

    c_2_o[(ticker + '_c_2_o')] = zscore(ret(D.o,D.c))
    h_2_o[(ticker + '_h_2_o')] = zscore(ret(D.o,D.h))
    l_2_o[(ticker + '_l_2_o')] = zscore(ret(D.o,D.l))
    c_2_h[(ticker + '_c_2_h')] = zscore(ret(D.h,D.c))
    h_2_l[(ticker + '_h_2_l')] = zscore(ret(D.h,D.l))
    c1_c0[(ticker + '_c1_c0')] = ret(D.c,D.c.shift(-1)).fillna(0) #Tommorows return 
    vol[(ticker + '_vol')] = zscore(D.v)

for f in os.listdir(datapath):
    filepath = os.path.join(datapath,f)
    if filepath.endswith('.csv'):
        make_inputs(filepath)

dates = c_2_o.index
pivot = list(c_2_o.columns) +list( h_2_o.columns) +list( l_2_o.columns) + list(c_2_h.columns) + list(h_2_l.columns) + list(c1_c0.columns) + list(vol.columns)
values = np.concatenate((c_2_o.values, h_2_o.values, l_2_o.values, c_2_h.values, h_2_l.values, c1_c0.values, vol.values), axis=1)
flat = pd.DataFrame(values, index=dates, columns=pivot)

I am getting error:
P.columns = new_ind

P = P.sort(axis=1) # Sort by columns

AttributeError Traceback (most recent call last)
in ()
1 P.columns = new_ind
----> 2 P = P.sort(axis=1) # Sort by columns

c:\python35\lib\site-packages\pandas\core\generic.py in getattr(self, name)
3079 if name in self._info_axis:
3080 return self[name]
-> 3081 return object.getattribute(self, name)
3082
3083 def setattr(self, name, value):

AttributeError: 'DataFrame' object has no attribute 'sort'

It seems sort is deprecated - now there is sort_values and sort_index.
Can you adapt code please?

Training the RNN¶

with tf.Graph().as_default():

model = RNNModel()

input_ = train[0]

target = train[1]

with tf.Session() as sess:

    init = tf.initialize_all_variables()

    sess.run([init])

    loss = 2000

    

    for e in range(NUM_EPOCHS):

        state = sess.run(model.zero_state)

        epoch_loss =0

        for batch in range(0,NUM_TRAIN_BATCHES):

            start = batch*BATCH_SIZE

            end = start + BATCH_SIZE 

            feed = {

                model.input_data:input_[start:end],

                model.target_data:target[start:end],

                model.dropout_prob:0.5,

                model.start_state:state

                        }

            _,loss,acc,state = sess.run(

                [

                    model.train_op,

                    model.loss,

                    model.accuracy,

                    model.end_state

                ]

                ,feed_dict=feed

            )

            epoch_loss+=loss

            

        print('step - {0} loss - {1} acc - {2}'.format((e),epoch_loss,acc))

    final_preds =np.array([])

    for batch in range(0,NUM_VAL_BATCHES):

            start = batch*BATCH_SIZE

            end = start + BATCH_SIZE 

            feed = {

                model.input_data:val[0][start:end],

                model.target_data:val[1][start:end],

                model.dropout_prob:1,

                model.start_state:state

                        }

            acc,preds,state = sess.run(

                [

                    model.accuracy,

                    model.predictions,

                    model.end_state

                ]

                ,feed_dict=feed

            )

            print(acc)

            assert len(preds) == BATCH_SIZE

            final_preds = np.concatenate((final_preds,preds),axis=0)

WARNING:tensorflow:<tensorflow.contrib.rnn.python.ops.rnn_cell.AttentionCellWrapper object at 0x0000021E0CFB02B0>: Using a concatenated state is slower and will soon be deprecated. Use state_is_tuple=True.

ValueError Traceback (most recent call last)
in ()
1 with tf.Graph().as_default():
----> 2 model = RNNModel()
3 input_ = train[0]
4 target = train[1]
5 with tf.Session() as sess:

in init(self)
44 scope.reuse_variables()
45
---> 46 output, state = self.gru_cell(inp, state)
47 states.append(state)
48 outputs.append(output)

c:\python35\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state, scope)
178 with vs.variable_scope(vs.get_variable_scope(),
179 custom_getter=self._rnn_get_variable):
--> 180 return super(RNNCell, self).call(inputs, state)
181
182 def _rnn_get_variable(self, getter, *args, **kwargs):

c:\python35\lib\site-packages\tensorflow\python\layers\base.py in call(self, inputs, *args, **kwargs)
439 # Check input assumptions set after layer building, e.g. input shape.
440 self._assert_input_compatibility(inputs)
--> 441 outputs = self.call(inputs, *args, **kwargs)
442
443 # Apply activity regularization.

c:\python35\lib\site-packages\tensorflow\contrib\rnn\python\ops\rnn_cell.py in call(self, inputs, state)
1111 input_size = inputs.get_shape().as_list()[1]
1112 inputs = _linear([inputs, attns], input_size, True)
-> 1113 lstm_output, new_state = self._cell(inputs, state)
1114 if self._state_is_tuple:
1115 new_state_cat = array_ops.concat(nest.flatten(new_state), 1)

c:\python35\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state, scope)
178 with vs.variable_scope(vs.get_variable_scope(),
179 custom_getter=self._rnn_get_variable):
--> 180 return super(RNNCell, self).call(inputs, state)
181
182 def _rnn_get_variable(self, getter, *args, **kwargs):

c:\python35\lib\site-packages\tensorflow\python\layers\base.py in call(self, inputs, *args, **kwargs)
439 # Check input assumptions set after layer building, e.g. input shape.
440 self._assert_input_compatibility(inputs)
--> 441 outputs = self.call(inputs, *args, **kwargs)
442
443 # Apply activity regularization.

c:\python35\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state)
914 [-1, cell.state_size])
915 cur_state_pos += cell.state_size
--> 916 cur_inp, new_state = cell(cur_inp, cur_state)
917 new_states.append(new_state)
918

c:\python35\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state, scope)
178 with vs.variable_scope(vs.get_variable_scope(),
179 custom_getter=self._rnn_get_variable):
--> 180 return super(RNNCell, self).call(inputs, state)
181
182 def _rnn_get_variable(self, getter, *args, **kwargs):

c:\python35\lib\site-packages\tensorflow\python\layers\base.py in call(self, inputs, *args, **kwargs)
439 # Check input assumptions set after layer building, e.g. input shape.
440 self._assert_input_compatibility(inputs)
--> 441 outputs = self.call(inputs, *args, **kwargs)
442
443 # Apply activity regularization.

c:\python35\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state)
293 value = math_ops.sigmoid(
294 _linear([inputs, state], 2 * self._num_units, True, bias_ones,
--> 295 self._kernel_initializer))
296 r, u = array_ops.split(value=value, num_or_size_splits=2, axis=1)
297 with vs.variable_scope("candidate"):

c:\python35\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in _linear(args, output_size, bias, bias_initializer, kernel_initializer)
1015 _WEIGHTS_VARIABLE_NAME, [total_arg_size, output_size],
1016 dtype=dtype,
-> 1017 initializer=kernel_initializer)
1018 if len(args) == 1:
1019 res = math_ops.matmul(args[0], weights)

c:\python35\lib\site-packages\tensorflow\python\ops\variable_scope.py in get_variable(name, shape, dtype, initializer, regularizer, trainable, collections, caching_device, partitioner, validate_shape, use_resource, custom_getter)
1063 collections=collections, caching_device=caching_device,
1064 partitioner=partitioner, validate_shape=validate_shape,
-> 1065 use_resource=use_resource, custom_getter=custom_getter)
1066 get_variable_or_local_docstring = (
1067 """%s

c:\python35\lib\site-packages\tensorflow\python\ops\variable_scope.py in get_variable(self, var_store, name, shape, dtype, initializer, regularizer, reuse, trainable, collections, caching_device, partitioner, validate_shape, use_resource, custom_getter)
960 collections=collections, caching_device=caching_device,
961 partitioner=partitioner, validate_shape=validate_shape,
--> 962 use_resource=use_resource, custom_getter=custom_getter)
963
964 def _get_partitioned_variable(self,

c:\python35\lib\site-packages\tensorflow\python\ops\variable_scope.py in get_variable(self, name, shape, dtype, initializer, regularizer, reuse, trainable, collections, caching_device, partitioner, validate_shape, use_resource, custom_getter)
358 reuse=reuse, trainable=trainable, collections=collections,
359 caching_device=caching_device, partitioner=partitioner,
--> 360 validate_shape=validate_shape, use_resource=use_resource)
361 else:
362 return _true_getter(

c:\python35\lib\site-packages\tensorflow\python\ops\variable_scope.py in wrapped_custom_getter(getter, *args, **kwargs)
1403 return custom_getter(
1404 functools.partial(old_getter, getter),
-> 1405 *args, **kwargs)
1406 return wrapped_custom_getter
1407

c:\python35\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in _rnn_get_variable(self, getter, *args, **kwargs)
181
182 def _rnn_get_variable(self, getter, *args, **kwargs):
--> 183 variable = getter(*args, **kwargs)
184 trainable = (variable in tf_variables.trainable_variables() or
185 (isinstance(variable, tf_variables.PartitionedVariable) and

c:\python35\lib\site-packages\tensorflow\python\ops\variable_scope.py in wrapped_custom_getter(getter, *args, **kwargs)
1403 return custom_getter(
1404 functools.partial(old_getter, getter),
-> 1405 *args, **kwargs)
1406 return wrapped_custom_getter
1407

c:\python35\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in _rnn_get_variable(self, getter, *args, **kwargs)
181
182 def _rnn_get_variable(self, getter, *args, **kwargs):
--> 183 variable = getter(*args, **kwargs)
184 trainable = (variable in tf_variables.trainable_variables() or
185 (isinstance(variable, tf_variables.PartitionedVariable) and

c:\python35\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in _rnn_get_variable(self, getter, *args, **kwargs)
181
182 def _rnn_get_variable(self, getter, *args, **kwargs):
--> 183 variable = getter(*args, **kwargs)
184 trainable = (variable in tf_variables.trainable_variables() or
185 (isinstance(variable, tf_variables.PartitionedVariable) and

c:\python35\lib\site-packages\tensorflow\python\ops\variable_scope.py in _true_getter(name, shape, dtype, initializer, regularizer, reuse, trainable, collections, caching_device, partitioner, validate_shape, use_resource)
350 trainable=trainable, collections=collections,
351 caching_device=caching_device, validate_shape=validate_shape,
--> 352 use_resource=use_resource)
353
354 if custom_getter is not None:

c:\python35\lib\site-packages\tensorflow\python\ops\variable_scope.py in _get_single_variable(self, name, shape, dtype, initializer, regularizer, partition_info, reuse, trainable, collections, caching_device, validate_shape, use_resource)
667 raise ValueError("Trying to share variable %s, but specified shape %s"
668 " and found shape %s." % (name, shape,
--> 669 found_var.get_shape()))
670 if not dtype.is_compatible_with(found_var.dtype):
671 dtype_str = dtype.name

ValueError: Trying to share variable rnn/attention_cell_wrapper/multi_rnn_cell/cell_0/gru_cell/gates/kernel, but specified shape (200, 200) and found shape (2428, 200).

I am getting the following error messages when I try and run the tensorflow portion of preparedata.ipynb. Not sure where to go from here.
with tf.Graph().as_default():
model = Model()
input_ = train[0]
target = train[1]
with tf.Session() as sess:
init = tf.initialize_all_variables()
sess.run([init])
epoch_loss =0
for e in range(NUM_EPOCHS):
if epoch_loss >0 and epoch_loss <1:
break
epoch_loss =0
for batch in range(0,NUM_TRAIN_BATCHES):

            start = batch*BATCH_SIZE
            end = start + BATCH_SIZE 
            feed = {
                model.input_data:input_[start:end],
                model.target_data:target[start:end],
                model.dropout_prob:0.9
                        }
            
            _,loss,acc = sess.run(
                [
                    model.train_op,
                    model.loss,
                    model.accuracy,
                ]
                ,feed_dict=feed
            )
            epoch_loss+=loss
        print('step - {0} loss - {1} acc - {2}'.format((1+batch+NUM_TRAIN_BATCHES*e),epoch_loss,acc))
            
    
    print('done training')
    final_preds =np.array([])
    final_probs =None
    for batch in range(0,NUM_VAL_BATCHES):
        
            start = batch*BATCH_SIZE
            end = start + BATCH_SIZE 
            feed = {
                model.input_data:val[0][start:end],
                model.target_data:val[1][start:end],
                model.dropout_prob:1
                        }
            
            acc,preds,probs = sess.run(
                [
                    model.accuracy,
                    model.predictions,
                    model.probs
                ]
                ,feed_dict=feed
            )
            print(acc)
            final_preds = np.concatenate((final_preds,preds),axis=0)
            if final_probs is None:
                final_probs = probs
            else:
                final_probs = np.concatenate((final_probs,probs),axis=0)
    prediction_conf = final_probs[np.argmax(final_probs,1)]

ValueError Traceback (most recent call last)
in ()
1 with tf.Graph().as_default():
----> 2 model = Model()
3 input_ = train[0]
4 target = train[1]
5 with tf.Session() as sess:

in init(self)
23 with tf.variable_scope("loss"):
24
---> 25 self.losses = tf.nn.sparse_softmax_cross_entropy_with_logits(self.logits,self.target_data)
26 mask = (1-tf.sign(1-self.target_data)) #Don't give credit for flat days
27 mask = tf.cast(mask,tf.float32)

/opt/intel/intelpython3/lib/python3.6/site-packages/tensorflow/python/ops/nn_ops.py in sparse_softmax_cross_entropy_with_logits(_sentinel, labels, logits, name)
2011 """
2012 _ensure_xent_args("sparse_softmax_cross_entropy_with_logits", _sentinel,
-> 2013 labels, logits)
2014
2015 # TODO(pcmurray) Raise an error when the label is not an index in

/opt/intel/intelpython3/lib/python3.6/site-packages/tensorflow/python/ops/nn_ops.py in _ensure_xent_args(name, sentinel, labels, logits)
1777 if sentinel is not None:
1778 raise ValueError("Only call %s with "
-> 1779 "named arguments (labels=..., logits=..., ...)" % name)
1780 if labels is None or logits is None:
1781 raise ValueError("Both labels and logits must be provided.")

ValueError: Only call sparse_softmax_cross_entropy_with_logits with named arguments (labels=..., logits=..., ...)

To work with latest TF package, you'll need to make certain modification to the last part namely RNN codes.
tf.nn.rnn_cell is no longer there and you'll need to use tf.contrib.rnn.*, function name is same.
And, I'm trying to use "'state_is_tuple=True" since it's recommended now, but failed...

BTW, why the RNN learning loss is 1000+ at last? Does that mean anything?

Howdy Tal,

Attempting to run this in Tensorflow v1.4, and running into the issue below when replacing tf.pack/unpack with tf.stack/unstack. Any thoughts on what's going on here? I attached full stack from the run in question.

Thanks!
Jim

WARNING:tensorflow:<tensorflow.contrib.rnn.python.ops.rnn_cell.AttentionCellWrapper object at 0x0000000048BDA588>: Using a concatenated state is slower and will soon be deprecated. Use state_is_tuple=True.
Tensor("Shape:0", shape=(3,), dtype=int32)
Tensor("Shape_1:0", shape=(3,), dtype=int32)

InvalidArgumentError Traceback (most recent call last)
C:\Anaconda3\lib\site-packages\tensorflow\python\framework\common_shapes.py in _call_cpp_shape_fn_impl(op, input_tensors_needed, input_tensors_as_shapes_needed, require_shape_fn)
685 graph_def_version, node_def_str, input_shapes, input_tensors,
--> 686 input_tensors_as_shapes, status)
687 except errors.InvalidArgumentError as err:

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\errors_impl.py in exit(self, type_arg, value_arg, traceback_arg)
472 compat.as_text(c_api.TF_Message(self.status.status)),
--> 473 c_api.TF_GetCode(self.status.status))
474 # Delete the underlying status object from memory otherwise it stays alive

WARNING:tensorflow:<tensorflow.contrib.rnn.python.ops.rnn_cell.AttentionCellWrapper object at 0x0000000048BDA588>: Using a concatenated state is slower and will soon be deprecated. Use state_is_tuple=True.
Tensor("Shape:0", shape=(3,), dtype=int32)
Tensor("Shape_1:0", shape=(3,), dtype=int32)

InvalidArgumentError Traceback (most recent call last)
C:\Anaconda3\lib\site-packages\tensorflow\python\framework\common_shapes.py in _call_cpp_shape_fn_impl(op, input_tensors_needed, input_tensors_as_shapes_needed, require_shape_fn)
685 graph_def_version, node_def_str, input_shapes, input_tensors,
--> 686 input_tensors_as_shapes, status)
687 except errors.InvalidArgumentError as err:

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\errors_impl.py in exit(self, type_arg, value_arg, traceback_arg)
472 compat.as_text(c_api.TF_Message(self.status.status)),
--> 473 c_api.TF_GetCode(self.status.status))
474 # Delete the underlying status object from memory otherwise it stays alive

InvalidArgumentError: Dimensions must be equal, but are 200 and 2428 for 'rnn/rnn/attention_cell_wrapper/attention_cell_wrapper_1/multi_rnn_cell/cell_0/cell_0/gru_cell/MatMul_2' (op: 'MatMul') with input shapes: [1,200], [2428,200].

During handling of the above exception, another exception occurred:

ValueError Traceback (most recent call last)
in ()
1 with tf.Graph().as_default():
----> 2 model = RNNModel()
3 input_ = train[0]
4 target = train[1]
5 with tf.Session() as sess:

in init(self)
41 scope.reuse_variables()
42
---> 43 output, state = self.gru_cell(inp, state)
44 states.append(state)
45 outputs.append(output)

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state, scope)
181 with vs.variable_scope(vs.get_variable_scope(),
182 custom_getter=self._rnn_get_variable):
--> 183 return super(RNNCell, self).call(inputs, state)
184
185 def _rnn_get_variable(self, getter, *args, **kwargs):

C:\Anaconda3\lib\site-packages\tensorflow\python\layers\base.py in call(self, inputs, *args, **kwargs)
573 if in_graph_mode:
574 self._assert_input_compatibility(inputs)
--> 575 outputs = self.call(inputs, *args, **kwargs)
576
577 if outputs is None:

C:\Anaconda3\lib\site-packages\tensorflow\contrib\rnn\python\ops\rnn_cell.py in call(self, inputs, state)
1117 self._linear1 = _Linear([inputs, attns], input_size, True)
1118 inputs = self._linear1([inputs, attns])
-> 1119 cell_output, new_state = self._cell(inputs, state)
1120 if self._state_is_tuple:
1121 new_state_cat = array_ops.concat(nest.flatten(new_state), 1)

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state, scope)
181 with vs.variable_scope(vs.get_variable_scope(),
182 custom_getter=self._rnn_get_variable):
--> 183 return super(RNNCell, self).call(inputs, state)
184
185 def _rnn_get_variable(self, getter, *args, **kwargs):

C:\Anaconda3\lib\site-packages\tensorflow\python\layers\base.py in call(self, inputs, *args, **kwargs)
573 if in_graph_mode:
574 self._assert_input_compatibility(inputs)
--> 575 outputs = self.call(inputs, *args, **kwargs)
576
577 if outputs is None:

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state)
1064 [-1, cell.state_size])
1065 cur_state_pos += cell.state_size
-> 1066 cur_inp, new_state = cell(cur_inp, cur_state)
1067 new_states.append(new_state)
1068

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state, scope)
181 with vs.variable_scope(vs.get_variable_scope(),
182 custom_getter=self._rnn_get_variable):
--> 183 return super(RNNCell, self).call(inputs, state)
184
185 def _rnn_get_variable(self, getter, *args, **kwargs):

C:\Anaconda3\lib\site-packages\tensorflow\python\layers\base.py in call(self, inputs, *args, **kwargs)
573 if in_graph_mode:
574 self._assert_input_compatibility(inputs)
--> 575 outputs = self.call(inputs, *args, **kwargs)
576
577 if outputs is None:

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state)
320 kernel_initializer=self._kernel_initializer)
321
--> 322 value = math_ops.sigmoid(self._gate_linear([inputs, state]))
323 r, u = array_ops.split(value=value, num_or_size_splits=2, axis=1)
324

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, args)
1187 res = math_ops.matmul(args[0], self._weights)
1188 else:
-> 1189 res = math_ops.matmul(array_ops.concat(args, 1), self._weights)
1190 if self._build_bias:
1191 res = nn_ops.bias_add(res, self._biases)

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\math_ops.py in matmul(a, b, transpose_a, transpose_b, adjoint_a, adjoint_b, a_is_sparse, b_is_sparse, name)
1889 else:
1890 return gen_math_ops._mat_mul(
-> 1891 a, b, transpose_a=transpose_a, transpose_b=transpose_b, name=name)
1892
1893

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\gen_math_ops.py in _mat_mul(a, b, transpose_a, transpose_b, name)
2434 _, _, _op = _op_def_lib._apply_op_helper(
2435 "MatMul", a=a, b=b, transpose_a=transpose_a, transpose_b=transpose_b,
-> 2436 name=name)
2437 _result = _op.outputs[:]
2438 _inputs_flat = _op.inputs

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\op_def_library.py in _apply_op_helper(self, op_type_name, name, **keywords)
785 op = g.create_op(op_type_name, inputs, output_types, name=scope,
786 input_types=input_types, attrs=attr_protos,
--> 787 op_def=op_def)
788 return output_structure, op_def.is_stateful, op
789

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\ops.py in create_op(self, op_type, inputs, dtypes, input_types, name, attrs, op_def, compute_shapes, compute_device)
2956 op_def=op_def)
2957 if compute_shapes:
-> 2958 set_shapes_for_outputs(ret)
2959 self._add_op(ret)
2960 self._record_op_seen_by_control_dependencies(ret)

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\ops.py in set_shapes_for_outputs(op)
2207 shape_func = _call_cpp_shape_fn_and_require_op
2208
-> 2209 shapes = shape_func(op)
2210 if shapes is None:
2211 raise RuntimeError(

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\ops.py in call_with_requiring(op)
2157
2158 def call_with_requiring(op):
-> 2159 return call_cpp_shape_fn(op, require_shape_fn=True)
2160
2161 _call_cpp_shape_fn_and_require_op = call_with_requiring

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\common_shapes.py in call_cpp_shape_fn(op, require_shape_fn)
625 res = _call_cpp_shape_fn_impl(op, input_tensors_needed,
626 input_tensors_as_shapes_needed,
--> 627 require_shape_fn)
628 if not isinstance(res, dict):
629 # Handles the case where _call_cpp_shape_fn_impl calls unknown_shape(op).

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\common_shapes.py in _call_cpp_shape_fn_impl(op, input_tensors_needed, input_tensors_as_shapes_needed, require_shape_fn)
689 missing_shape_fn = True
690 else:
--> 691 raise ValueError(err.message)
692
693 if missing_shape_fn:

ValueError: Dimensions must be equal, but are 200 and 2428 for 'rnn/rnn/attention_cell_wrapper/attention_cell_wrapper_1/multi_rnn_cell/cell_0/cell_0/gru_cell/MatMul_2' (op: 'MatMul') with input shapes: [1,200], [2428,200].

During handling of the above exception, another exception occurred:

ValueError Traceback (most recent call last)
in ()
1 with tf.Graph().as_default():
----> 2 model = RNNModel()
3 input_ = train[0]
4 target = train[1]
5 with tf.Session() as sess:

in init(self)
41 scope.reuse_variables()
42
---> 43 output, state = self.gru_cell(inp, state)
44 states.append(state)
45 outputs.append(output)

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state, scope)
181 with vs.variable_scope(vs.get_variable_scope(),
182 custom_getter=self._rnn_get_variable):
--> 183 return super(RNNCell, self).call(inputs, state)
184
185 def _rnn_get_variable(self, getter, *args, **kwargs):

C:\Anaconda3\lib\site-packages\tensorflow\python\layers\base.py in call(self, inputs, *args, **kwargs)
573 if in_graph_mode:
574 self._assert_input_compatibility(inputs)
--> 575 outputs = self.call(inputs, *args, **kwargs)
576
577 if outputs is None:

C:\Anaconda3\lib\site-packages\tensorflow\contrib\rnn\python\ops\rnn_cell.py in call(self, inputs, state)
1117 self._linear1 = _Linear([inputs, attns], input_size, True)
1118 inputs = self._linear1([inputs, attns])
-> 1119 cell_output, new_state = self._cell(inputs, state)
1120 if self._state_is_tuple:
1121 new_state_cat = array_ops.concat(nest.flatten(new_state), 1)

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state, scope)
181 with vs.variable_scope(vs.get_variable_scope(),
182 custom_getter=self._rnn_get_variable):
--> 183 return super(RNNCell, self).call(inputs, state)
184
185 def _rnn_get_variable(self, getter, *args, **kwargs):

C:\Anaconda3\lib\site-packages\tensorflow\python\layers\base.py in call(self, inputs, *args, **kwargs)
573 if in_graph_mode:
574 self._assert_input_compatibility(inputs)
--> 575 outputs = self.call(inputs, *args, **kwargs)
576
577 if outputs is None:

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state)
1064 [-1, cell.state_size])
1065 cur_state_pos += cell.state_size
-> 1066 cur_inp, new_state = cell(cur_inp, cur_state)
1067 new_states.append(new_state)
1068

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state, scope)
181 with vs.variable_scope(vs.get_variable_scope(),
182 custom_getter=self._rnn_get_variable):
--> 183 return super(RNNCell, self).call(inputs, state)
184
185 def _rnn_get_variable(self, getter, *args, **kwargs):

C:\Anaconda3\lib\site-packages\tensorflow\python\layers\base.py in call(self, inputs, *args, **kwargs)
573 if in_graph_mode:
574 self._assert_input_compatibility(inputs)
--> 575 outputs = self.call(inputs, *args, **kwargs)
576
577 if outputs is None:

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, inputs, state)
320 kernel_initializer=self._kernel_initializer)
321
--> 322 value = math_ops.sigmoid(self._gate_linear([inputs, state]))
323 r, u = array_ops.split(value=value, num_or_size_splits=2, axis=1)
324

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\rnn_cell_impl.py in call(self, args)
1187 res = math_ops.matmul(args[0], self._weights)
1188 else:
-> 1189 res = math_ops.matmul(array_ops.concat(args, 1), self._weights)
1190 if self._build_bias:
1191 res = nn_ops.bias_add(res, self._biases)

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\math_ops.py in matmul(a, b, transpose_a, transpose_b, adjoint_a, adjoint_b, a_is_sparse, b_is_sparse, name)
1889 else:
1890 return gen_math_ops._mat_mul(
-> 1891 a, b, transpose_a=transpose_a, transpose_b=transpose_b, name=name)
1892
1893

C:\Anaconda3\lib\site-packages\tensorflow\python\ops\gen_math_ops.py in _mat_mul(a, b, transpose_a, transpose_b, name)
2434 _, _, _op = _op_def_lib._apply_op_helper(
2435 "MatMul", a=a, b=b, transpose_a=transpose_a, transpose_b=transpose_b,
-> 2436 name=name)
2437 _result = _op.outputs[:]
2438 _inputs_flat = _op.inputs

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\op_def_library.py in _apply_op_helper(self, op_type_name, name, **keywords)
785 op = g.create_op(op_type_name, inputs, output_types, name=scope,
786 input_types=input_types, attrs=attr_protos,
--> 787 op_def=op_def)
788 return output_structure, op_def.is_stateful, op
789

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\ops.py in create_op(self, op_type, inputs, dtypes, input_types, name, attrs, op_def, compute_shapes, compute_device)
2956 op_def=op_def)
2957 if compute_shapes:
-> 2958 set_shapes_for_outputs(ret)
2959 self._add_op(ret)
2960 self._record_op_seen_by_control_dependencies(ret)

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\ops.py in set_shapes_for_outputs(op)
2207 shape_func = _call_cpp_shape_fn_and_require_op
2208
-> 2209 shapes = shape_func(op)
2210 if shapes is None:
2211 raise RuntimeError(

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\ops.py in call_with_requiring(op)
2157
2158 def call_with_requiring(op):
-> 2159 return call_cpp_shape_fn(op, require_shape_fn=True)
2160
2161 _call_cpp_shape_fn_and_require_op = call_with_requiring

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\common_shapes.py in call_cpp_shape_fn(op, require_shape_fn)
625 res = _call_cpp_shape_fn_impl(op, input_tensors_needed,
626 input_tensors_as_shapes_needed,
--> 627 require_shape_fn)
628 if not isinstance(res, dict):
629 # Handles the case where _call_cpp_shape_fn_impl calls unknown_shape(op).

C:\Anaconda3\lib\site-packages\tensorflow\python\framework\common_shapes.py in _call_cpp_shape_fn_impl(op, input_tensors_needed, input_tensors_as_shapes_needed, require_shape_fn)
689 missing_shape_fn = True
690 else:
--> 691 raise ValueError(err.message)
692
693 if missing_shape_fn:

ValueError: Dimensions must be equal, but are 200 and 2428 for 'rnn/rnn/attention_cell_wrapper/attention_cell_wrapper_1/multi_rnn_cell/cell_0/cell_0/gru_cell/MatMul_2' (op: 'MatMul') with input shapes: [1,200], [2428,200].

In
In [23]: TotalReturn = ((1-exp(TargetDF)).sum(1))/num_stocks

Normally to go from log return to simple return, we do:
R = exp(r) – 1
But you have done opposite
R = 1 - exp(r)

Is there something I misunderstood?

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
ERROR

('self.logits = ', <tf.Tensor 'ff/fully_connected_2/BiasAdd:0' shape=(?, 11) dtype=float32>)
('self.target_data', <tf.Tensor 'Placeholder_1:0' shape=(?,) dtype=int32>)
Traceback (most recent call last):
File "./preparedata-manual-upgraded.py", line 204, in
model = Model()
File "./preparedata-manual-upgraded.py", line 187, in init
self.losses = tf.nn.sparse_softmax_cross_entropy_with_logits(labels=self.logits,logits=self.target_data)
File "/home/steven/Practical-DataScience/DataScience/local/lib/python2.7/site-packages/tensorflow/python/ops/nn_ops.py", line 1686, in sparse_softmax_cross_entropy_with_logits
(labels_static_shape.ndims, logits.get_shape().ndims))
ValueError: Rank mismatch: Rank of labels (received 2) should equal rank of logits minus 1 (received 1)

CODE IN QUESTION

class Model():
def init(self):
global_step = tf.contrib.framework.get_or_create_global_step()
self.input_data = tf.placeholder(dtype=tf.float32,shape=[None,num_features])
self.target_data = tf.placeholder(dtype=tf.int32,shape=[None])
self.dropout_prob = tf.placeholder(dtype=tf.float32,shape=[])
with tf.variable_scope("ff"):
droped_input = tf.nn.dropout(self.input_data,keep_prob=self.dropout_prob)

        layer_1 = tf.contrib.layers.fully_connected(
            num_outputs=hidden_1_size,
            inputs=droped_input,
        )
        layer_2 = tf.contrib.layers.fully_connected(
            num_outputs=hidden_2_size,
            inputs=layer_1,
        )
        self.logits = tf.contrib.layers.fully_connected(
            num_outputs=num_classes,
            activation_fn =None,
            inputs=layer_2,
        )
    with tf.variable_scope("loss"):
        print ("self.logits = ",self.logits) 
        print ("self.target_data", self.target_data)

exit()

        self.losses = tf.nn.sparse_softmax_cross_entropy_with_logits(labels=self.logits,logits=self.target_data)
        mask = (1-tf.sign(1-self.target_data)) #Don't give credit for flat days
        mask = tf.cast(mask,tf.float32)
        self.loss = tf.reduce_sum(self.losses)
    
    with tf.name_scope("train"):
      opt = tf.train.AdamOptimizer(lr)
      gvs = opt.compute_gradients(self.loss)
      self.train_op = opt.apply_gradients(gvs, global_step=global_step)
    
    with tf.name_scope("predictions"):
        self.probs = tf.nn.softmax(self.logits)
        self.predictions = tf.argmax(self.probs, 1)
        correct_pred = tf.cast(tf.equal(self.predictions, tf.cast(self.target_data,tf.int64)),tf.float64)
        self.accuracy = tf.reduce_mean(correct_pred)

PRINTED OUTPUT OF VARIABLES BEFORE ENTERING THE FUNCTION

[[ 2 3 11 6 1 7 7 3 3 4 5]
[ 2 3 8 7 8 7 6 2 2 2 3]
[ 1 4 9 5 2 13 5 11 5 3 2]
[ 1 6 7 8 5 15 6 1 7 4 2]
[ 1 3 6 2 3 9 10 5 7 4 0]
[ 0 5 11 3 3 6 6 4 5 6 2]
[ 1 3 15 3 3 12 12 1 4 2 4]
[ 0 4 8 3 3 8 12 2 10 3 0]
[ 0 2 16 6 3 9 12 2 3 0 1]
[ 0 7 11 5 2 7 10 4 4 4 2]
[ 3 3 10 6 6 9 6 4 1 4 0]]
[[91 37 75]
[76 35 89]
[92 30 75]]

RELEVANT TENSOR FLOW DOC

https://www.tensorflow.org/api_docs/python/tf/nn/sparse_softmax_cross_entropy_with_logits