I've looked for a similar situation but I'm getting this problem after implementing what I thought should be the fix. I'm making a LSTM model to predict a solar system's power output using a data set from kaggle. The function is

def power_prediction_model():
    df = pd.read_csv('powerData/power_plant_generation_data_limited.csv')
    df = df.rename(columns={'DATE_TIME': 'Date'})

    df = df[['Date', 'AC_POWER']]

    df['Date'] = pd.to_datetime(df['Date'])

    data = df.filter(['AC_POWER'])

    dataset = data.values

    training_data_len = math.ceil(len(dataset) * .8)

    # Scale the data
    scaler = MinMaxScaler(feature_range=(0, 1))
    scaled_data = scaler.fit_transform(dataset)

    # Create the training data set
    train_data = scaled_data[0:training_data_len, :]

    # Split the data into x_train and y_train data sets
    x_train = []  # independent training variables
    y_train = []  # dependant or target variable

    for i in range(1500, len(train_data)):
        x_train.append(train_data[i - 1500:i, 0])
        y_train.append(train_data[i, 0])

    # convert to numpy arrays
    x_train = np.array(x_train)
    y_train = np.array(y_train)

    # reshape the data to 3D
    x_train = np.reshape(x_train, (x_train.shape[0], x_train.shape[1], 1))

    # build the LSTM model
    model = Sequential()
    model.add(LSTM(50, return_sequences=True, input_shape=(x_train[1], 1)))
    model.add(LSTM(50, return_sequences=False))
    model.add(Dense(25))
    model.add(Dense(1))

with the error coming from the line model.add(LSTM(50, return_sequences=True, input_shape=(x_train[1], 1))) The array being passed to the model is converted to a numpy array so I'm stuck on where the TypeError is coming from.

CodePudding user response：

x_train[1] need to be x_train.shape[1]