使用ARIMA预测股价时遇ufunc 'isnan'类型兼容错误求助

错误原因 #

核心问题是循环调用arima_prediction时，传入的是股票代码字符串，而非DataFrame中对应的股价数据列。

看这段循环代码：

for stock in chosen_stocks:
    arima_prediction(stock)

这里的stock是字符串（比如'CTSH'），到了arima_prediction函数里，train_data, test_data = stock[3:int(len(dataframe) * 0.5)], stock[int(len(dataframe) * 0.5):]实际是对字符串做切片，得到的是单个字符组成的列表，并非数值型时间序列数据。ARIMA模型接收非数值类型数据后，内部调用np.isnan时自然报错，因为字符串无法进行空值判断。

修复步骤 #

只需修改循环部分，传入DataFrame中对应的列即可：

dataframe = get_data()
for stock in chosen_stocks:
    arima_prediction(dataframe[stock])  # 改为传入dataframe的对应列

额外优化建议 #

1. 函数内的len(dataframe)应改为len(stock_series)，因为函数接收的是单只股票的Series，用原DataFrame的长度会导致逻辑错误（比如单只股票数据长度与原DataFrame不一致时）：

def arima_prediction(stock_series):
    # 用len(stock_series)代替len(dataframe)
    train_data, test_data = stock_series[3:int(len(stock_series) * 0.5)], stock_series[int(len(stock_series) * 0.5):]
    # 后续代码保持不变

2. 可在函数开头添加类型检查，避免同类错误：

def arima_prediction(stock_series):
    if not isinstance(stock_series, pd.Series):
        raise ValueError("请传入Pandas Series类型的股价数据")
    # 后续代码

原问题信息 #

报错信息 #

None if faux_endog else np.any(np.isnan(self.endog))) TypeError: ufunc 'isnan' not supported for the input types, and the inputs could not be safely coerced to any supported types according to the casting rule 'safe'

完整代码 #

# Imports
import os
import yfinance as yf
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import io
from PIL import Image
import statsmodels.api as sm
from statsmodels.tsa.arima.model import ARIMA
from sklearn.metrics import mean_squared_error, mean_absolute_error
import math
from sklearn.preprocessing import MinMaxScaler
import tensorflow as tf

# Chosen stocks from NASDAQ-100
chosen_stocks = ['CTSH', 'BKNG', 'REGN', 'MSFT']

def get_data():
    # Get list of tickers
    tickers = open("dataset/nasdaq_100_tickers.txt", "r")
    data = tickers.read().splitlines()

    # Check if the data has already been downloaded, drop NaN values
    if os.path.exists('dataframe.csv'):
        dataframe = pd.read_csv('dataframe.csv', index_col="Date", parse_dates=True).dropna()
    else:
        # Download Close data from Yahoo Finance
        data = yf.download(tickers=data, period='1y', interval='1d')['Close']
        data.to_csv('dataframe.csv')
        # Convert array to Pandas dataframe, drop NaN values
        complete_data = data.dropna()
        dataframe = pd.DataFrame(complete_data)
    dataframe.drop(['GEHC'], axis=1, inplace=True) # Dropping GEHC because it contains NULL values

    return dataframe


def arima_prediction(stock):
    train_data, test_data = stock[3:int(len(dataframe) * 0.5)], stock[int(len(dataframe) * 0.5):]
    train_arima = train_data
    test_arima = test_data

    history = [x for x in train_arima]
    y = test_arima
    predictions = list()
    model = ARIMA(history, order=(1, 1, 0))
    model_fit = model.fit()
    forecast = model_fit.forecast()[0]
    predictions.append(forecast)
    history.append(y[0])

    for i in range(1, len(y)):
        # Predict
        model = ARIMA(history, order=(1, 1, 0))
        model_fit = model.fit()
        forecast = model_fit.forecast()[0]
        # Invert transformed prediction
        predictions.append(forecast)
        # Observation
        observation = y[i]
        history.append(observation)

    # Report performance
    mean_squared = mean_squared_error(y, predictions)
    print('Mean Squared Error: ' + str(mean_squared))
    mean_absolute = mean_absolute_error(y, predictions)
    print('Mean Absolute Error: ' + str(mean_absolute))
    root_mean_squared = math.sqrt(mean_squared_error(y, predictions))
    print('Root Mean Squared Error: ' + str(root_mean_squared))

dataframe = get_data()
for stock in chosen_stocks:
    arima_prediction(stock)

DataFrame样例 #

AAPL        ABNB  ...         ZM          ZS
Date                                ...                       
2022-12-15  136.500000   90.610001  ...  70.199997  117.169998
2022-12-16  134.509995   89.570000  ...  69.860001  114.209999
2022-12-19  132.369995   85.930000  ...  69.089996  112.269997
2022-12-20  132.300003   87.620003  ...  68.559998  113.540001
2022-12-21  135.449997   87.070000  ...  69.930000  112.769997
...                ...         ...  ...        ...         ...
2023-11-28  190.399994  127.559998  ...  67.529999  193.850006
2023-11-29  189.369995  126.480003  ...  67.949997  199.839996
2023-11-30  189.949997  126.339996  ...  67.830002  197.529999
2023-12-01  191.240005  135.020004  ...  70.290001  198.029999
2023-12-04  188.669998  134.539993  ...  67.720001  197.919998

完整报错回溯 #

Traceback (most recent call last):
  File "C:/Users/xxx/source/repos/Project/main.py", line 370, in <module>
    arima_prediction(stock)
  File "C:/Users/xxx/source/repos/Project/main.py", line 217, in arima_prediction
    model = ARIMA(history, order=(1, 1, 0))
  File "C:\Users\xxx\source\repos\Project\venv\lib\site-packages\statsmodels\tsa\arima\model.py", line 158, in __init__
    self._spec_arima = SARIMAXSpecification(
  File "C:\Users\xxx\source\repos\Project\venv\lib\site-packages\statsmodels\tsa\arima\specification.py", line 458, in __init__
    None if faux_endog else np.any(np.isnan(self.endog)))
TypeError: ufunc 'isnan' not supported for the input types, and the inputs could not be safely coerced to any supported types according to the casting rule ''safe''

Process finished with exit code 1

内容的提问来源于stack exchange，提问作者custardbun