Python数据分析pands中的Merge与join

Posted 2021-03-09 奔跑的金鱼

目录

• 1.Merge
  • 1.1 简单关联:left_on与right_on
  • 1.2 使用how参数:指定连接方式
  • 1.3 right_index与right_index
  • 1.4 sort参数：排序
• 2.join

1.Merge

Pandas具有全功能的,高性能内存中连接操作,与关系型数据库中的连接操作类似。
语法：

pd.merge(left, right, how=\'inner\', on=None, left_on=None, right_on=None,
         left_index=False, right_index=False, sort=True,
         suffixes=(\'_x\', \'_y\'), copy=True, indicator=False)

1.1 简单关联:left_on与right_on

下面是Merge的一些实战案例：
（1）创建测试数组

import pandas as pd
df1 = pd.DataFrame({\'studentNo\':[\'A01\',\'A02\',\'A03\',\'A04\'],
                   \'studentName\':[\'Jack\',\'Lucy\',\'Marry\',\'Tom\'],
                   \'studentAge\':[18,19,21,17],
                   \'classNo\':[\'class01\',\'class02\',\'class03\',\'class02\']
                   })
df2 = pd.DataFrame({\'classNo\':[\'class01\',\'class02\',\'class03\',\'class04\'],
                     \'className\':[\'火箭1班\',\'火箭2班\',\'火箭3班\',\'火箭4班\']
                     })

# 学生信息
print(df1)

# 班级信息
print(df2)

（2）使用Merge，找出每个学生对应的班级名字

print(pd.merge(df1, df2, on="classNo")) # 这里的df1类似关系型数据库的主表，df2对应
# 如果存在多个键关联，则on = [\'key1\',\'key2\']

输出结果：

1.2 使用how参数:指定连接方式

# 1.取交集：默认取交集
print(\'--------参数how = \\\'inner\\\'的结果--------------\')  
print(pd.merge(df1, df2, on="classNo",how=\'inner\'))  # 类似内连接
# 2.取并集，数据缺失则为NaN
print(\'--------参数how = \\\'outer\\\'的结果--------------\')
print(pd.merge(df1, df2, on="classNo",how=\'outer\')) # 类似全连接

# 3.左连接
print(\'--------参数how = \\\'left\\\'的结果--------------\')
print(pd.merge(df1, df2, on="classNo",how=\'left\')) 

# 4.右连接
print(\'--------参数how = \\\'right\\\'的结果--------------\')
print(pd.merge(df1, df2, on="classNo",how=\'right\'))  

1.3 right_index与right_index

import pandas as pd

df1 = pd.DataFrame({\'left_key\':list(\'abcd\'),
                    \'data1\':range(4)})

df2 = pd.DataFrame({\'right_key\':list(\'acef\'),
                    \'data2\':range(4)})
# df1 以‘lkey’为键，df2以\'rkey\'为键
print(pd.merge(df1,df2,left_on=\'left_key\',right_on=\'right_key\'))         

df1 = pd.DataFrame({\'key\':list(\'abcdfeg\'),
                   \'data1\':range(7)})
df2 = pd.DataFrame({\'data2\':range(1,6)},
                  index = list(\'abcde\'))
print(\'----------df1-------------\')
print(df1)
print(\'----------df2-------------\')
print(df2)
print(\'----------df1,df2,参数:left_on=\\\'key\\\',right_index=True-------------\')
print(pd.merge(df1,df2,left_on=\'key\',right_index=True)) 
# df1采用key列值作为关联数据，df2采用index作为关联数据
# left_index：为True时，第一个df以index为键，默认False
# right_index：为True时，第二个df以index为键，默认False

# 所以left_on, right_on, left_index, right_index可以相互组合

1.4 sort参数：排序

import pandas as pd
# 参数sort 
df1 = pd.DataFrame({\'key\':list(\'baecd\'),
                   \'data1\':[10,2,4,7,5]})
df2 = pd.DataFrame({\'key\':list(\'abc\'),
                    \'data2\':[11,2,33]})
print(\'-------------df1-------------\')
print(df1)
print(\'-------------df2-------------\')
print(df2)

y1 = pd.merge(df1,df2, on = \'key\', how = \'outer\')
y2 = pd.merge(df1,df2, on = \'key\', sort=True, how = \'outer\')
print(\'-------------df1与df2全连接的结果-------------\')
print(y1)
print(\'-------------df1与df2全连接并根据连接键排序的结果-------------\')
print(y2)
# sort：按照字典顺序通过 连接键 对结果DataFrame进行排序。默认为False，设置为False会大幅提高性能
print(\'-------------df1与df2全连接的结果根据data1排序-------------\')
# 也可以连接完毕后直接使用DataFrame的排序方法
print(y2.sort_values(\'data1\'))

2.join

import pandas as pd
# pd.join 通过索引连接
left = pd.DataFrame({\'A\':[\'A01\',\'A02\',\'A03\'],
                     \'B\':[\'B01\',\'B02\',\'B03\']
                    },index=[\'k1\',\'k2\',\'k3\'])

right = pd.DataFrame({\'C\':[\'C01\',\'C02\',\'C03\',\'C04\'],
                     \'D\':[\'D01\',\'D02\',\'D03\',\'D04\']
                    },index=[\'k1\',\'k2\',\'k3\',\'k4\'])

print(left,\'-----------left-----------\\n\')
print(left,\'-----------right-----------\\n\')
print(left.join(right),\'-----------left.join(right)-----------\\n\')
print(left.join(right,how=\'outer\'),\'-----------left.join(right)-----------\\n\')
# 上述语句等价于 pd.merge(left, right, left_index=True, right_index=True, how=\'outer\')

df1 = pd.DataFrame({\'key\':list(\'bbacde\'),
                  \'data1\':[1,3,5,7,2,4]})

df2 = pd.DataFrame({\'key\':list(\'abc\'),
                   \'data2\':[11,2,15]
                   })

print(\'-----------df1-----------\')
print(df1)

print(\'-----------df2-----------\')
print(df2)

# 指定参数suffixes
print(\'-\' * 50)
print(pd.merge(df1, df2, left_index=True, right_index=True, suffixes=(\'_1\', \'_2\')))  

# 用df1中的索引与df2[\'data2\']中的索引连接
print(\'-\' * 50)
# print(df2[\'data2\'])
print(df1.join(df2[\'data2\']))

left = pd.DataFrame({\'A\': [\'A0\', \'A1\', \'A2\', \'A3\'],
                     \'B\': [\'B0\', \'B1\', \'B2\', \'B3\'],
                     \'key\': [\'K0\', \'K1\', \'K0\', \'K1\']})
right = pd.DataFrame({\'C\': [\'C0\', \'C1\'],
                      \'D\': [\'D0\', \'D1\']},
                     index=[\'K0\', \'K1\'])
print(left)
print(right)
print(left.join(right, on = \'key\'))
# 等价于pd.merge(left, right, left_on=\'key\', right_index=True, how=\'left\', sort=False);
# left的‘key’和right的index

输出结果：