Error when checking input: expected lstm_1_input to have 3 dimensions

Question

I have an x_test data:

x_test.shape
Out[11]: (13096, 30)

x_test.size
Out[16]: 392880

When I launch the prediction it return an error:

ValueError: Error when checking input: expected lstm_1_input to have 3 dimensions, but got array with shape (13096, 30)

My code is very simple, I'm trying to run the predict function:

test_df= pd.read_csv("Path_data")
model = load_model.("path_model")
Xnew = np.array(test_df)

# make a prediction

predictions = model.predict_classes(Xnew)

Model summary:

Model: "sequential_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
lstm_1 (LSTM)                (None, 50, 100)           50400     
_________________________________________________________________
dropout_1 (Dropout)          (None, 50, 100)           0         
_________________________________________________________________
lstm_2 (LSTM)                (None, 50)                30200     
_________________________________________________________________
dropout_2 (Dropout)          (None, 50)                0         
_________________________________________________________________
dense_1 (Dense)              (None, 1)                 51        
=================================================================
Total params: 80,651
Trainable params: 80,651
Non-trainable params: 0

Someone please can tell me how can I resolve this issue ? Thank you

Answer 1

This is dimension mismatch, the model expect to have a data record itself as an array of arrays, ie, if you have an image i, it expects, np.array([i]), Try :

x_train = np.array([ [i] for i in x_train])

This will increment the dimension by 1

Answer 2

Your variable x_test should have the input dimensions required by the LSTM layer which is a 3D tensor with shape [batch, timesteps, feature]