I am trying to run a robust (to outliers) regression in Python using sm.RLM and while computing the p-values of the estimated coefficients I want to use HAC (heteroskedasticity and autocorrelation consistent) (Newey-West) standard errors. Can someone please let me know how I can do this.
I have tried to do something similar to what you would do for OLS: sm.OLS(y,x).fit(cov_type=“HAC”, cov_keds={“maxlags”:365}) but it does not work with sm.RLM. (365 since I am using daily data) And I have also tried using sw.cov_hac but couldn’t get it working