Как устранить ошибку при проверке ввода модели: ожидалось, что convolution2d_input_9 будет иметь 4 измерения, но получит массив с формой

Question

Я использую существующую модель Keras, которую сделал мой сокурсник, и пытаюсь внедрить CNN, но я продолжаю получать эту ошибку

ValueError: Error when checking model input: expected convolution2d_input_9 to have 4 dimensions, but got array with shape (3938, 4, 42)

Вот лог

    runfile('C:/Users/a_phi/Downloads/hand-pred-20190324T215735Z-001/hand-pred/model_lstmAshwin-train.py', wdir='C:/Users/a_phi/Downloads/hand-pred-20190324T215735Z-001/hand-pred')
Reloaded modules: myFileHandler
x_train (3938, 4, 42)
y_train (3938, 6)
saving scaler object to Scaler.sav
Compilation Time :  0.026177644729614258
____________________________________________________________________________________________________
Layer (type)                     Output Shape          Param #     Connected to                     
====================================================================================================
convolution2d_24 (Convolution2D) (None, 26, 26, 32)    320         convolution2d_input_11[0][0]     
____________________________________________________________________________________________________
activation_10 (Activation)       (None, 26, 26, 32)    0           convolution2d_24[0][0]           
____________________________________________________________________________________________________
convolution2d_25 (Convolution2D) (None, 24, 24, 64)    18496       activation_10[0][0]              
____________________________________________________________________________________________________
maxpooling2d_12 (MaxPooling2D)   (None, 12, 12, 64)    0           convolution2d_25[0][0]           
____________________________________________________________________________________________________
dropout_22 (Dropout)             (None, 12, 12, 64)    0           maxpooling2d_12[0][0]            
____________________________________________________________________________________________________
flatten_10 (Flatten)             (None, 9216)          0           dropout_22[0][0]                 
____________________________________________________________________________________________________
dense_20 (Dense)                 (None, 128)           1179776     flatten_10[0][0]                 
____________________________________________________________________________________________________
dropout_23 (Dropout)             (None, 128)           0           dense_20[0][0]                   
____________________________________________________________________________________________________
dense_21 (Dense)                 (None, 10)            1290        dropout_23[0][0]                 
====================================================================================================
Total params: 1,199,882
Trainable params: 1,199,882
Non-trainable params: 0
____________________________________________________________________________________________________
Now to train the model using the fit() method
Traceback (most recent call last):

  File "<ipython-input-19-d0bd53cbab62>", line 1, in <module>
    runfile('C:/Users/a_phi/Downloads/hand-pred-20190324T215735Z-001/hand-pred/model_lstmAshwin-train.py', wdir='C:/Users/a_phi/Downloads/hand-pred-20190324T215735Z-001/hand-pred')

  File "C:\Users\a_phi\Anaconda3\envs\cpr_lstm07\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 668, in runfile
    execfile(filename, namespace)

  File "C:\Users\a_phi\Anaconda3\envs\cpr_lstm07\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 108, in execfile
    exec(compile(f.read(), filename, 'exec'), namespace)

  File "C:/Users/a_phi/Downloads/hand-pred-20190324T215735Z-001/hand-pred/model_lstmAshwin-train.py", line 627, in <module>
    verbose=2)

  File "C:\Users\a_phi\Anaconda3\envs\cpr_lstm07\lib\site-packages\keras\models.py", line 672, in fit
    initial_epoch=initial_epoch)

  File "C:\Users\a_phi\Anaconda3\envs\cpr_lstm07\lib\site-packages\keras\engine\training.py", line 1116, in fit
    batch_size=batch_size)

  File "C:\Users\a_phi\Anaconda3\envs\cpr_lstm07\lib\site-packages\keras\engine\training.py", line 1029, in _standardize_user_data
    exception_prefix='model input')

  File "C:\Users\a_phi\Anaconda3\envs\cpr_lstm07\lib\site-packages\keras\engine\training.py", line 112, in standardize_input_data
    str(array.shape))

ValueError: Error when checking model input: expected convolution2d_input_11 to have 4 dimensions, but got array with shape (3938, 4, 42)

Я незнаком с Керасом и не уверен, как это исправить.

def build_model2(dataIn, timeWindow, layerOut, outputs, activation, 
        optimizer):

        layerOut = 84
        layerOutExpand = 100
        dataIn = 42
        timeWindow = 20
        outputs = 6  


        # Takes feature data matrix row size as input = 42
        # Returns vector of size 20 (timeWindow) as output
        num_classes = 10
        model = Sequential()

        model.add(Convolution2D(32, 3, 3, input_shape=(28, 28, 1)))
        model.add(Activation('relu'))
        model.add(Conv2D(64,3,3))
        model.add(MaxPooling2D(pool_size=(2, 2)))
        model.add(Dropout(0.25))
        model.add(Flatten())
        model.add(Dense(128, activation='relu'))
        model.add(Dropout(0.5))
        model.add(Dense(num_classes, activation='softmax'))

        # Compile the model and time how long it takes
        start = time.time()
        #optimizer = SGD(lr=0.3, momentum=0.9)
        model.compile(loss="mse", optimizer=optimizer, metrics=[coeffDetermination, 'accuracy'])
        print("Compilation Time : ", time.time() - start)

        # Let's see the model details too, and create an image file of the 
        # network structure
        model.summary() 

    return model

Другие решения, которые я видел, я не мог понять или заставить их работать.

Я не знаком с Keras / TensorFlow, поэтому любая помощь будет принята.