Keras ValueError：无法重塑大小数组

Question

我在使用 Keras 进行预测任务时遇到了一个我无法理解的错误。

这是我的代码：

import numpy as np
import pandas as pd
from sklearn.preprocessing import MinMaxScaler
from keras.models import Sequential
from keras.layers import LSTM, Dense, Dropout
import matplotlib.pyplot as plt
import finpy_tse as fpy

# Define the stock symbol and time period
stock_symbol = 'وغدیر'
start_date = '1400-01-01'
end_date = '1402-01-01'

# Download stock price data from the Tehran Stock Exchange using finpy_tse
stock_data = fpy.Get_Price_History(stock_symbol, start_date, end_date)

# Print the fetched data
print("Fetched data from the site:")
print(stock_data)

data = pd.DataFrame(stock_data)

# Download stock price data from the Tehran Stock Exchange using finpy_tse
stock_data = fpy.Get_Price_History(stock_symbol, start_date, end_date)
data = pd.DataFrame(stock_data)

# Print columns to inspect the structure of the DataFrame
print(data.columns)

# Feature engineering: Creating additional features if needed
# Assuming the 'Close' or 'Final' and 'Volume' columns are present in the DataFrame
if 'Close' in data.columns and 'Volume' in data.columns:
    data['DailyReturn'] = data['Close'].pct_change()
    data['VolumeChange'] = data['Volume'].pct_change()

    # Drop missing values
    data = data.dropna()

    # Check if the dataset is empty after dropping missing values
    if data.empty:
        print("Empty dataset after dropping missing values. Unable to proceed.")
    else:
        # Use multiple features for training
        features = ['DailyReturn', 'VolumeChange']
        dataset = data[features].values

        # Check if the dataset is empty after feature engineering
        if dataset.shape[0] == 0:
            print("Empty dataset after feature engineering. Unable to proceed.")
        else:
            # Normalize the data only if the dataset is not empty
            scaler = MinMaxScaler(feature_range=(0, 1))
            dataset_scaled = scaler.fit_transform(dataset)

            # Continue with the rest of your code...

            # Create a time series dataset
            def create_dataset(dataset, time_steps=1):
                X, y = [], []
                for i in range(len(dataset) - time_steps):
                    a = dataset[i:(i + time_steps), :]
                    X.append(a)
                    y.append(dataset[i + time_steps, 0])  # Predicting DailyReturn
                return np.array(X), np.array(y)

            # Define the time steps (you can adjust this parameter)
            time_steps = 10

            # Create the time series dataset
            X, y = create_dataset(dataset_scaled, time_steps)

            # Print shape and size for debugging
            print("Before reshaping - X shape:", X.shape, "X size:", X.size)

            # Reshape data for LSTM model [samples, time steps, features]
            X = np.reshape(X, (1, X.shape[0], X.shape[1]))

            # Build the LSTM model with dropout for better generalization
            model = Sequential()
            model.add(LSTM(units=100, return_sequences=True, input_shape=(X.shape[1], X.shape[2])))
            model.add(Dropout(0.2))
            model.add(LSTM(units=100, return_sequences=True))
            model.add(Dropout(0.2))
            model.add(LSTM(units=50, return_sequences=False))
            model.add(Dropout(0.2))
            model.add(Dense(units=1))
            model.compile(optimizer='adam', loss='mean_squared_error')

            # Train the model
            model.fit(X, y, epochs=10, batch_size=32, validation_split=0.1)

            # Test the model on new data
            test_data = dataset_scaled[-time_steps:]
            test_data = np.reshape(test_data, (1, time_steps, len(features)))
            predicted_price = model.predict(test_data)

            # Inverse transform and reshape the predicted_price
            predicted_price = scaler.inverse_transform(np.reshape(predicted_price, (1, time_steps, len(features))))
            print("Shapes for debugging:")
            print("test_data shape:", test_data.shape)
            print("predicted_price shape:", predicted_price.shape)

            # Reshape and inverse transform
            predicted_price = scaler.inverse_transform(np.reshape(predicted_price, (time_steps, len(features))))

            # Assuming you are interested in the first column of the prediction
            predicted_price = predicted_price[0, :, 0]

            # Visualize the results
            plt.plot(data['Close'].values, label='Actual Stock Price')
            plt.plot(np.arange(len(data['Close']), len(data['Close']) + time_steps), predicted_price, marker='o', color='red',
                     label='Predicted Stock Price')
            plt.xlabel('Days')
            plt.ylabel('Stock Price')
            plt.title(f'Stock Price Prediction for {stock_symbol} using LSTM')
            plt.legend()
            plt.show()
else:
    print("The 'Close' or 'Volume' columns are not present in the DataFrame.")
    print("Please check your stock price data or provide a DataFrame with these columns.")

错误是：

回溯（最近一次调用最后一次）：文件 “...:/Users/...../PycharmProjects/untitled1/.....py”，第 76 行，在 X = np.reshape(X, (1, X.shape[0], X.shape[1])) File "<array_function insides>"，第 180 行，在 reshape 文件中 “....:\Users.....\PycharmProjects\untitled1 env\lib\site-packages 尤姆普