我试图找到一些有关如何正确解释OLS回归结果的信息:
model = ols("rating ~ gender_F + gender_M + genre + gender_F*genre + gender_M*genre", data = data).fit()
model_summary = model.summary()
print(model_summary)
我得到以下输出:
OLS Regression Results
==============================================================================
Dep. Variable: rating R-squared: 0.000
Model: OLS Adj. R-squared: -0.000
Method: Least Squares F-statistic: 0.3757
Date: Tue, 15 Sep 2020 Prob (F-statistic): 1.00
Time: 17:48:10 Log-Likelihood: -1.5372e+05
No. Observations: 100000 AIC: 3.075e+05
Df Residuals: 99963 BIC: 3.079e+05
Df Model: 36
Covariance Type: nonrobust
=================================================================================================
coef std err t P>|t| [0.025 0.975]
-------------------------------------------------------------------------------------------------
Intercept -8.39e+11 1.26e+12 -0.663 0.507 -3.32e+12 1.64e+12
genre[T.Adventure] -1.94e+11 1.31e+11 -1.481 0.139 -4.51e+11 6.28e+10
genre[T.Animation] -2.204e+11 3.09e+11 -0.714 0.475 -8.25e+11 3.85e+11
genre[T.Childrens] 1.299e+11 3.06e+11 0.424 0.671 -4.7e+11 7.3e+11
genre[T.Comedy] -1.17e+11 2.24e+11 -0.523 0.601 -5.55e+11 3.21e+11
genre[T.Crime] 1.739e+11 3.52e+11 0.493 0.622 -5.17e+11 8.65e+11
genre[T.Documentary] 9.41e+11 7.97e+11 1.181 0.238 -6.2e+11 2.5e+12
genre[T.Drama] 1.276e+11 7.25e+11 0.176 0.860 -1.29e+12 1.55e+12
........
gender_F 8.39e+11 1.26e+12 0.663 0.507 -1.64e+12 3.32e+12
gender_F:genre[T.Adventure] 1.94e+11 1.31e+11 1.481 0.139 -6.28e+10 4.51e+11
gender_F:genre[T.Animation] 2.204e+11 3.09e+11 0.714 0.475 -3.85e+11 8.25e+11
gender_F:genre[T.Childrens] -1.299e+11 3.06e+11 -0.424 0.671 -7.3e+11 4.7e+11
gender_F:genre[T.Comedy] 1.17e+11 2.24e+11 0.523 0.601 -3.21e+11 5.55e+11
gender_F:genre[T.Crime] -1.739e+11 3.52e+11 -0.493 0.622 -8.65e+11 5.17e+11
gender_F:genre[T.Documentary] -9.41e+11 7.97e+11 -1.181 0.238 -2.5e+12 6.2e+11
gender_F:genre[T.Drama] -1.276e+11 7.25e+11 -0.176 0.860 -1.55e+12 1.29e+12
gender_F:genre[T.Fantasy] -2.866e+10 6.69e+10 -0.429 0.668 -1.6e+11 1.02e+11
gender_F:genre[T.Film-Noir] -6.506e+11 1.02e+12 -0.636 0.525 -2.66e+12 1.36e+12
......
我想知道如何解释该系数?我们决定将该系数视为有效的临界值是多少?分析中还应包含哪些其他内容?解释P> | t |是p值(在我的情况下没有显着关系)?