Applying operation to a pandas multi index dataframe subgroup

[Nuovoq]

Aint it always the way - five minutes of searching later I found a solution:

dfForex['last_close'] = dfForex.groupby(['<ticker>']).shift()['<close>']

Result:

dfForex[['<close>', 'last_close']]
Out[3]: 
                     <close>  last_close
<date>     <ticker>                     
2019-07-01 AUDEUR     0.6170         NaN
           AUDGBP     0.5508         NaN
           AUDHKD     5.4413         NaN
           AUDJPY    75.4780         NaN
           AUDUSD     0.6965         NaN
2019-07-03 AUDEUR     0.6229      0.6170
           AUDGBP     0.5587      0.5508
           AUDHKD     5.4818      5.4413
           AUDJPY    75.7710     75.4780
           AUDUSD     0.7031      0.6965
2019-07-04 AUDEUR     0.6224      0.6229
           AUDGBP     0.5583      0.5587
           AUDHKD     5.4687      5.4818
           AUDJPY    75.7310     75.7710
           AUDUSD     0.7024      0.7031