pandas.core.groupby.GroupBy.mean

GroupBy.mean(*args, **kwargs)[source]

Compute mean of groups, excluding missing values

For multiple groupings, the result index will be a MultiIndex