np.median(WeibullAFTFitter().fit(data).predict_median(data)) vs WeibullAFTFitter().fit(data).median_survival_time_

[jack-fletch]

I am newer to Survival Analysis. I am trying to wrap my head around why np.median(WeibullAFTFitter().fit(data).predict_median(data)) returns a different (usually higher) value than WeibullAFTFitter().fit(data).median_survival_time_. Any help or direction anyone could provide would be greatly appreciated :)