new.txt

滚动回归部分考虑更换为标准库statsmodels内的RollingOLS类，对应版本：0.11

安装地址：
https://github.com/statsmodels/statsmodels/releases/tag/v0.11.0rc2

实例代码：

import time
import numpy as np
stock_count = 4000
days = 3000
window = 30
from multiprocessing import Pool

def rolling_reg(d):
    X = d[:, :-1]
    Y = d[:, -1]
    from statsmodels.regression.rolling import RollingOLS
    result = RollingOLS(Y, X, window).fit()
    return result.params


if __name__ == '__main__':
  st = time.time()
  with Pool() as p:
     results = p.map(rolling_reg, np.random.rand(stock_count, days, 11))
  print('滚动：%ss' % (time.time() - st))