python dataframe where

在dataframe实现类似SQL的where过滤

# -*- coding: UTF-8 -*-
import pandas as pd
df = pd.DataFrame([{'col1':'a', 'col2':1}, {'col1':'b', 'col2':2}, {'col1':'c', 'col2':3}])

# 单条件
print df.loc[df['col1'] == 'a']
print df.loc[df['col1'] != 'a']
print df.loc[df['col2'] > 2]
print df.loc[df['col1'].isin(['a', 'b'])]

# 多条件
print df.loc[df['col1'] == 'c'].loc[df['col2'] > 1]

输出结果:

  col1  col2
0    a     1
  col1  col2
1    b     2
2    c     3
  col1  col2
2    c     3
  col1  col2
0    a     1
1    b     2
  col1  col2
2    c     3

dataframe空值过滤

import pandas as pd
df1 = pd.DataFrame([{'col1':'a', 'col2':1}, {'col1':'b', 'col2':2}])
df2 = pd.DataFrame([{'col1':'a', 'col3':11}, {'col1':'c', 'col3':33}])

data = pd.merge(left=df1, right=df2, how='left', left_on='col1', right_on='col1')
print data
print data.loc[data['col3'].isnull()]
print data.loc[data['col3'].notnull()]

输出结果:

  col1  col2  col3
0    a     1    11
1    b     2   NaN
  col1  col2  col3
1    b     2   NaN
  col1  col2  col3
0    a     1    11