pandas 带聚合函数的groupby
pandas groupby with aggregate function
我有这个代码:
october_data_grouped = october_data.groupby(["Department", "Headcount", "Day"]).agg({'User ID': 'nunique'})
october_data_grouped.unstack(fill_value=0)
这给出了以下输出:
有没有一种方法可以计算出每个单独的条目除以人数[例如对于行帐户,第 1 天为 17% (6/35),第 4 天为 25% (9/35) 等]
使用MultiIndex.get_level_values
with DataFrame.div
:
#for avoid MultiIndex in columns
october_data_grouped = (october_data.groupby(["Department", "Headcount", "Day"])['User ID']
.nunique()
.unstack(fill_value=0))
idx = october_data_grouped.index.get_level_values('Headcount')
october_data_grouped = october_data_grouped.div(idx, axis=0)
我有这个代码:
october_data_grouped = october_data.groupby(["Department", "Headcount", "Day"]).agg({'User ID': 'nunique'})
october_data_grouped.unstack(fill_value=0)
这给出了以下输出:
有没有一种方法可以计算出每个单独的条目除以人数[例如对于行帐户,第 1 天为 17% (6/35),第 4 天为 25% (9/35) 等]
使用MultiIndex.get_level_values
with DataFrame.div
:
#for avoid MultiIndex in columns
october_data_grouped = (october_data.groupby(["Department", "Headcount", "Day"])['User ID']
.nunique()
.unstack(fill_value=0))
idx = october_data_grouped.index.get_level_values('Headcount')
october_data_grouped = october_data_grouped.div(idx, axis=0)