Creating New Columns In Sublevel Of Multindex Pandas Columns

I have a multindex column. Higher level is some humans, sublevel is some measures. I would like to create some new columns which are derivatives of measures (eg. rolling mean).

Solution 1:

First select columns by xs, apply rolling and create MultiIndex, last join to original:

df = xf.xs('bezel', axis=1, level=1).rolling(window=2).mean()
df.columns = [df.columns, ['roll2'] * len(df.columns)]

Another solution with rename:

df = (xf.xs('bezel', axis=1, level=1, drop_level=False).rolling(window=2).mean()
        .rename(columns={'bezel':'roll2'}))

---

print(df)humanmikedavemattroll2roll2roll22018-01-31       NaNNaNNaN2018-02-28  0.4392970.7565300.4076062018-03-31  0.4325130.4366600.4303932018-04-30  0.2587360.4696100.8509962018-05-31  0.2788690.6988220.561285

---

xf=xf.join(df)print(xf)humanmikedave\measurespindragbezelspindragbezel2018-01-31  0.8110300.1145350.3265790.5977810.1940640.6597952018-02-28  0.7749710.4008880.5520160.3855390.5823510.8532662018-03-31  0.7944270.6534280.3130100.9965140.5249990.0200552018-04-30  0.3074180.1314510.2044620.0493460.1988780.9191652018-05-31  0.1963740.4215940.3532760.2440240.9309920.478479humanmattmikedave\measurespindragbezelrollroll2roll22018-01-31  0.7693080.6579630.691395NaNNaNNaN2018-02-28  0.5648840.0268640.1238180.4392970.4392970.7565302018-03-31  0.7554400.6984430.7369670.4325130.4325130.4366602018-04-30  0.7829080.9190640.9650250.2587360.2587360.4696102018-05-31  0.4140850.3397710.1575450.2788690.2788690.698822humanmattmeasureroll22018-01-31       NaN2018-02-28  0.4076062018-03-31  0.4303932018-04-30  0.8509962018-05-31  0.561285

Last if encessary sorting MultiIndex:

xf = xf.join(df).sort_index(axis=1)