给出两个pandas数据帧:
df1 = pd.read_csv(file1, names=['col1','col2','col3'])
df2 = pd.read_csv(file2, names=['col1','col2','col3'])
我想删除df2中df1中不存在col1
或col2
(或两者)值的所有行。
执行以下操作:
df2 = df2[(df2['col1'] in set(df1['col1'])) & (df2['col2'] in set(df1['col2']))]
的产率:
TypeError:'系列'对象是可变的,因此它们不能被散列
答案 0 :(得分:2)
我认为你可以试试isin
:
df2 = df2[(df2['col1'].isin(df1['col1'])) & (df2['col2'].isin(df1['col2']))]
df1 = pd.DataFrame({'col1':[1,2,3,3],
'col2':[4,5,6,2],
'col3':[7,8,9,5]})
print (df1)
col1 col2 col3
0 1 4 7
1 2 5 8
2 3 6 9
3 3 2 5
df2 = pd.DataFrame({'col1':[1,2,3,5],
'col2':[4,7,4,1],
'col3':[7,8,9,1]})
print (df2)
col1 col2 col3
0 1 4 7
1 2 7 8
2 3 4 9
3 5 1 1
df2 = df2[(df2['col1'].isin(df1['col1'])) & (df2['col2'].isin(df1['col2'].unique()))]
print (df2)
col1 col2 col3
0 1 4 7
2 3 4 9
另一个解决方案是merge
,因为内联接(how='inner'
)是默认设置,但它仅适用于DataFrames
中具有相同位置的值:
print (pd.merge(df1, df2))
col1 col2 col3
0 1 4 7