Pandas newey west weight is always 0

In the newey_west function in pandas.stats.math, the weight variable is always gonna be 0, since max_lags and lag are both integers and lag < max_lags+1. So the newey-west lag has no effect on the Xeps. Is that right?

    Xeps = np.dot(m.T, m)
    for lag in range(1, max_lags + 1):
        auto_cov = np.dot(m[:-lag].T, m[lag:])
        weight = lag / (max_lags + 1)
        if nw_overlap:
            weight = 0
        bb = auto_cov + auto_cov.T
        dd = (1 - weight) * bb
        Xeps += dd
    Xeps *= nobs / (nobs - df)

Comment From: jorisvandenbossche

I am not familiar with the Newey West implementation, but note that both the pandas.stats.plm as pandas.stats.ols modules that use pandas.stats.math.newey_west are deprecated, and we refer to statsmodels for such functionality. Can you check if you can find all functionality you need in statsmodels?

Comment From: TomAugspurger

@ZZZWENG see here for how to do this. Specifically, you want cov_type='HAC',cov_kwds={'maxlags':1} when fitting the model.

Comment From: ZZWENG

Thanks!