Pandas 如何合并数据框

Pandas是一个开源的Python库，提供高性能的数据操作和分析工具，使用其强大的数据结构。在Pandas中，数据框是一种二维数据结构，即数据以行和列的表格形式对齐。

在本文中，我们将看到如何在Python中合并数据框。我们将使用merge()方法。以下是语法：

dataframe.merge(right, how, on, left_on, right_on, left_index, right_index, sort, suffixes, copy, indicator, validate)

这里,

参数	值	描述
right		要合并的DataFrame或Series
how	‘left’ ‘right’ ‘outer’ ‘inner’: 默认 ‘cross’	合并方式
on	字符串列表	进行合并的级别
left_on	字符串列表	进行合并的DataFrame左侧的级别
right_on	字符串列表	进行合并的DataFrame右侧的级别
left_index	True False	是否将左侧DataFrame的索引作为连接键
right_index	True False	是否将右侧DataFrame的索引作为连接键
sort	True False	是否按连接键对DataFrame进行排序
suffixes	列表	用于重叠列的字符串列表
copy	True False

使用merge()方法合并数据帧，使用右侧数据帧的键

要合并数据帧，我们将使用merge()方法。how参数的右值仅使用右侧数据帧的键，类似于SQL的右外连接。

示例

import pandas as pd

# Create Dictionaries
dct1 = {'Player':['Steve','David'], 'Age':[29, 25,]}
dct2 = {'Player':['Steve','Kane'], 'Age':[31, 27]}

# Create DataFrame from Dictionary elements using pandas.dataframe()
df1 = pd.DataFrame(dct1)
df2 = pd.DataFrame(dct2)
print("DataFrame1 = \n",df1)
print("\nDataFrame2 = \n",df2)

# Combining DataFrames using the merge() method
res = df1.merge(df2, how='right')
print("\nCombined DataFrames = \n",res)

输出

DataFrame1 = 
   Player  Age
0  Steve   29
1  David   25

DataFrame2 = 
   Player  Age
0  Steve   31
1   Kane   27
Combined DataFrames = 
   Player  Age
0  Steve   31
1   Kane   27

使用merge()方法从左侧数据框中使用键合并数据框

为了合并数据框，我们将使用merge()方法。how参数的左值仅使用左侧数据框中的键，类似于SQL的左外连接。

示例

import pandas as pd

# Create Dictionaries
dct1 = {'Player':['Steve','David'], 'Age':[29, 25,]}
dct2 = {'Player':['Steve','Kane'], 'Age':[31, 27]}

# Create DataFrame from Dictionary elements using pandas.dataframe()
df1 = pd.DataFrame(dct1)
df2 = pd.DataFrame(dct2)
print("DataFrame1 = \n",df1)
print("\nDataFrame2 = \n",df2)

# Combining DataFrames using the merge() method

# The how parameter is set to left
res = df1.merge(df2, how='left')
print("\nCombined DataFrames = \n",res)

输出

DataFrame1 = 
   Player  Age
0  Steve   29
1  David   25

DataFrame2 = 
   Player  Age
0  Steve   31
1   Kane   27
Combined DataFrames = 
   Player  Age
0  Steve   29
1  David   25

从两个数据框中合并具有键的并集

要合并数据框，我们将使用merge()方法。how参数的外部取值使用两个框架的键的并集，类似于SQL的全外连接。

示例

import pandas as pd

# Create Dictionaries
dct1 = {'Player':['Steve','David'], 'Age':[29, 25,]}
dct2 = {'Player':['Steve','Kane'], 'Age':[31, 27]}

# Create DataFrame from Dictionary elements using pandas.dataframe()
df1 = pd.DataFrame(dct1)
df2 = pd.DataFrame(dct2)
print("DataFrame1 = \n",df1)
print("\nDataFrame2 = \n",df2)

# Combining DataFrames using the merge() method
# The how parameter is set to outer i.e.
res = df1.merge(df2, how='outer')
print("\nCombined DataFrames = \n",res)

输出

DataFrame1 = 
   Player  Age
0  Steve   29
1  David   25

DataFrame2 = 
   Player  Age
0  Steve   31
1   Kane   27
Combined DataFrames = 
   Player  Age
0  Steve   29
1  David   25
2  Steve   31
3   Kane   27

从两个数据帧中使用键的交集合并数据帧

要合并数据帧，我们将使用merge()方法。how参数的内部值使用两个数据帧的键的交集，类似于SQL的内连接。

示例

import pandas as pd

# Create Dictionaries
dct1 = {'Player':['Steve','David'], 'Age':[29, 25,]}
dct2 = {'Player':['Steve','Kane'], 'Age':[31, 27]}

# Create DataFrame from Dictionary elements using pandas.dataframe()
df1 = pd.DataFrame(dct1)
df2 = pd.DataFrame(dct2)
print("DataFrame1 = \n",df1)
print("\nDataFrame2 = \n",df2)

# Combining DataFrames using the merge() method
# The how parameter is set to inner
res = df1.merge(df2, how='inner')
print("\nCombined DataFrames = \n",res)

输出

DataFrame1 = 
   Player  Age
0  Steve   29
1  David   25

DataFrame2 = 
   Player  Age
0  Steve   31
1   Kane   27
Combined DataFrames = 
Empty DataFrame
Columns: [Player, Age]
Index: []

从两个数据框中进行笛卡尔乘积合并

要合并数据框，我们将使用merge()方法。how参数的交叉值会从两个框架中创建笛卡尔乘积：

示例

import pandas as pd

# Create Dictionaries
dct1 = {'Player':['Steve','David'], 'Age':[29, 25,]}
dct2 = {'Player':['Steve','Kane'], 'Age':[31, 27]}

# Create DataFrame from Dictionary elements using pandas.dataframe()
df1 = pd.DataFrame(dct1)
df2 = pd.DataFrame(dct2)
print("DataFrame1 = \n",df1)
print("\nDataFrame2 = \n",df2)

# Combining DataFrames using the merge() method
# The how parameter is set to cross i.e. cartesian product
res = df1.merge(df2, how='cross')
print("\nCombined DataFrames = \n",res)

输出

DataFrame1 = 
   Player  Age
0  Steve   29
1  David   25
DataFrame2 = 
    Player  Age
0   Steve   31
1   Kane    27Combined DataFrames = 
Player_x  Age_x Player_y  Age_y
0    Steve     29    Steve     31
1    Steve     29     Kane     27
2    David     25    Steve     31
3    David     25     Kane     27