问如何回避这个警告？

不用copy如何回避这个警告 [SettingWithCopyWarning: A value is trying to be set on a copy of a slice from a DataFrame. Try using .loc[row_indexer,col_indexer] = value ]

不一个一个要素算也可以用loc回避吗？

dfg1 = [df[df['name'] == na1] for na1 in df['name'].unique()]

for dfa in dfg:

dfg2 = [df[df['val'] == na2] for na2 in df['val '].unique()]

for dfb in dfg:

dfb['a1']=dfb['b3']-dfb['b1']