First I fit to the large scales (\(k<0.02\)) from the Walther P(k)

Compared to the Fit to the BOSS dataset: