eliorc/medium

Code related to blog posts on my Medium page

medium's Issues

Execution failed

I am trying to replicate the work of node2vec in pycharm , but got the following error. not sure why this is happening. Tried with less number of teams as well to reduce the data complexity , but nothing changed!
Can someone help me resolve this issue?

Plotting error

I got an error here...

plt.plot(malstm_trained.history['acc'])
plt.plot(malstm_trained.history['val_acc'])

---

KeyError Traceback (most recent call last)
in
1 # Plot accuracy
----> 2 plt.plot(malstm_trained.history['acc'])
3 plt.plot(malstm_trained.history['val_acc'])
4 plt.title('Model Accuracy')
5 plt.ylabel('Accuracy')

KeyError: 'acc'

how to predict?

How to get the GRU layer output?

Hi,
I should get the GRU layer output vector for use other distance method such as cosine. My code as follows:
left_input = Input(shape=(input_shape,), dtype='int32', name="left_input_layer") right_input = Input(shape=(input_shape,), dtype='int32', name="right_input_layer") embedding_layer = Embedding(max_features+1, embed_dims, name="embedding_layer") encoded_left = embedding_layer(left_input) encoded_right = embedding_layer(right_input) left_output = GRU(128, name="rnn_layer_left")(encoded_left) right_output = GRU(128, name="rnn_layer_right")(encoded_right) malstm_distance = Lambda(function=lambda x: exponent_neg_manhattan_distance(x[0], x[1]), output_shape=lambda x: (x[0][0], 1))([left_output, right_output])

and get GRU layer output code:
intermediate_layer_model = Model(inputs=model.input, outputs=model.get_layer(layer_name).output) intermediate_output = intermediate_layer_model.predict([idx1, idx2])

However, the different sentence gets the same vector, but when I predict the probability it's OK. Can you give me some suggestions?

Object float has no lens()

`max_seq_length = max(train_df.question1.map(lambda x: len(x)).max(),
train_df.question2.map(lambda x: len(x)).max(),
test_df.question1.map(lambda x: len(x)).max(),
test_df.question2.map(lambda x: len(x)).max())

validation_size = 40000
training_size = len(train_df) - validation_size

X = train_df[questions_cols]
Y = train_df['is_duplicate']

X_train, X_validation, Y_train, Y_validation = train_test_split(X, Y, test_size=validation_size)

X_train = {'left': X_train.question1, 'right': X_train.question2}
X_validation = {'left': X_validation.question1, 'right': X_validation.question2}
X_test = {'left': test_df.question1, 'right': test_df.question2}

Y_train = Y_train.values
Y_validation = Y_validation.values

for dataset, side in itertools.product([X_train, X_validation], ['left', 'right']):
dataset[side] = pad_sequences(dataset[side], maxlen=max_seq_length)

assert X_train['left'].shape == X_train['right'].shape
assert len(X_train['left']) == len(Y_train)`

Error I am getting is

TypeError Traceback (most recent call last)
in
----> 1 max_seq_length = max(train_df.question1.map(lambda x: len(x)).max(),
2 train_df.question2.map(lambda x: len(x)).max(),
3 test_df.question1.map(lambda x: len(x)).max(),
4 test_df.question2.map(lambda x: len(x)).max())
5

~.conda\envs\gpuversion\lib\site-packages\pandas\core\series.py in map(self, arg, na_action)
3628 dtype: object
3629 """
-> 3630 new_values = super()._map_values(arg, na_action=na_action)
3631 return self._constructor(new_values, index=self.index).finalize(self)
3632

~.conda\envs\gpuversion\lib\site-packages\pandas\core\base.py in _map_values(self, mapper, na_action)
1143
1144 # mapper is a function
-> 1145 new_values = map_f(values, mapper)
1146
1147 return new_values

pandas_libs\lib.pyx in pandas._libs.lib.map_infer()

in (x)
----> 1 max_seq_length = max(train_df.question1.map(lambda x: len(x)).max(),
2 train_df.question2.map(lambda x: len(x)).max(),
3 test_df.question1.map(lambda x: len(x)).max(),
4 test_df.question2.map(lambda x: len(x)).max())
5

TypeError: object of type 'float' has no len()

`

but when I use len(str(x)) to solve this issue I get another error of ValueError: sequences must be a list of iterables. Found non-iterable: nan

the variable X_test is never used in whole program ?

Is the variable X_test never consumed in whole program ? What is this variable's purpose ?

Error while calculating the distance

I couldn't import Merge using the line "from keras.layers import LSTM, Embedding, Input, Merge".
So I tried changing the code to "from keras.layers import LSTM, Embedding, Input, merge" which helped in executing this line but when I was executing the below line of code I got the below error. Please help

TypeError Traceback (most recent call last)
in ()
12 # Calculates the distance as defined by the MaLSTM model
13 malstm_distance = merge(mode=lambda x: exponent_neg_manhattan_distance(x[0], x[1]),
---> 14 output_shape=lambda x: (x[0][0], 1))([left_output, right_output])
15
16 # Combine all of the above in a Model

TypeError: 'module' object is not callable

node2vec: Embeddings for Graph Data

hi ,

while executing your code of node2vec following error was occurred, how to resolve it kindly suggest me

Traceback (most recent call last):

File "", line 15, in
node2vec = Node2Vec(graph, dimensions=20, walk_length=16, num_walks=100, workers=2)

File "C:\Users\pshrikant\AppData\Local\Continuum\anaconda3\lib\site-packages\node2vec\node2vec.py", line 71, in init
self.walks = self._generate_walks()

File "C:\Users\pshrikant\AppData\Local\Continuum\anaconda3\lib\site-packages\node2vec\node2vec.py", line 158, in _generate_walks
in enumerate(num_walks_lists, 1))

File "C:\Users\pshrikant\AppData\Local\Continuum\anaconda3\lib\site-packages\joblib\parallel.py", line 934, in call
self.retrieve()

File "C:\Users\pshrikant\AppData\Local\Continuum\anaconda3\lib\site-packages\joblib\parallel.py", line 833, in retrieve
self._output.extend(job.get(timeout=self.timeout))

TypeError: 'NoneType' object is not iterable