White.py

import matplotlib.pyplot as plt
import statsmodels.api as sm
from scipy import stats
import pandas as pd
import numpy as np


def whitetest(session: dict):
    dependent = session["dependent"]
    independent = session["independent"]
    data = pd.read_csv(session["filename"]).dropna()
    y = data[dependent]
    x = data[independent]
    coef = session["ans"]["coefficient"]
    tmp = np.array(coef[1:]).T * x
    # print(tmp.values)
    yhat = []
    for i in tmp.values:
        yhat.append(sum(i) + coef[0])
    u = np.array(yhat) - y
    u2 = u**2
    X = sm.add_constant(data[independent])
    white = sm.OLS(u2, X)
    white = white.fit()
    R_square = np.round(white.rsquared, 4)
    chi = R_square * len(x)
    pvalue = stats.chi2.sf(chi, len(coef))
    chi1 = stats.chi2.cdf(0.99, len(coef))
    chi2 = stats.chi2.cdf(0.95, len(coef))
    chi3 = stats.chi2.cdf(0.99, len(coef))
    if pvalue > chi1:
        return "<h2>White Test</h2><p>chi({})={}***</p><p>99% heteroscedasticity</p>".format(
            len(coef), chi)
    if pvalue > chi2:
        return "<h2>White Test</h2><p>chi2({})={}**</p><p>95% heteroscedasticity</p>".format(
            len(coef), chi)
    if pvalue > chi3:
        return "<h2>White Test</h2><p>chi2({})={}*</p><p>90% heteroscedasticity</p>".format(
            len(coef), chi)
    return "<center><h2>White Test</h2><p>chi({})={}</p></center>".format(
        len(coef), chi)


if __name__ == '__main__':
    print(
        whitetest({
            "filename":
            "/Users/dcy/code/asdfghjkl/sourceCode/tk/daily_Ashare.csv",
            "dependent": "open",
            "independent": ["high", "low"],
            "ans": {
                "coefficient": [1, 2, 3]
            }
        }))