跳转至

使用 Pandas apply()方法返回多列

原文:https://www.geesforgeks.org/return-multi-columns-use-pandas-apply-method/

传递给pants.apply()的对象是系列对象，其索引是数据框的索引(轴=0)或数据框的列(轴=1)。默认情况下(result_type=None)，最终的返回类型是从应用的函数的返回类型推断出来的。否则，它取决于 result_type 参数。

语法: DataFrame.apply(func，axis=0，broadcast=None，raw=False，reduce=None，result_type=None，args=()，**kwds)

以下是一些描述使用pandas.DataFrame.apply()的程序

例 1:

# Program to illustrate the use of 
# pandas.DataFrame.apply() method

# Importing required Libraries
import pandas
import numpy

# Creating dataframe
dataFrame = pandas.DataFrame([[4, 9], ] * 3, columns =['A', 'B'])
print('Data Frame:')
display(dataFrame)

# Using pandas.DataFrame.apply() on the data frame
print('Returning multiple columns from Pandas apply()')
dataFrame.apply(numpy.sqrt)

输出:

pandas-apply-1 使用 numpy 通用函数(在这种情况下与 numpy.sqrt(dataFrame)相同)。

例 2:

# Program to illustrate the use of 
# pandas.DataFrame.apply() method

# Importing required Libraries
import pandas
import numpy

# Creating dataframe
dataFrame = pandas.DataFrame([[4, 9], ] * 3, columns =['A', 'B'])
print('Data Frame:')
display(dataFrame)

# Using pandas.DataFrame.apply() on the data frame
print('Returning multiple columns from Pandas apply()')
dataFrame.apply(numpy.sum, axis = 0)

输出:

pandas-apply-2 在列上使用缩减函数。

例 3:

# Program to illustrate the use of 
# pandas.DataFrame.apply() method

# Importing required Libraries
import pandas
import numpy

# Creating dataframe
dataFrame = pandas.DataFrame([[4, 9], ] * 3, columns =['A', 'B'])
print('Data Frame:')
display(dataFrame)

# Using pandas.DataFrame.apply() on the data frame
print('Returning multiple columns from Pandas apply()')
dataFrame.apply(numpy.sum, axis = 1)

输出:

pandas-apply-3

在行上使用缩减函数。

例 4:

# Program to illustrate the use of 
# pandas.DataFrame.apply() method

# Importing required Libraries
import pandas
import numpy

# Creating dataframe
dataFrame = pandas.DataFrame([[4, 9], ] * 3, columns =['A', 'B'])
print('Data Frame:')
display(dataFrame)

# Using pandas.DataFrame.apply() on the data frame
print('Returning multiple columns from Pandas apply()')
dataFrame.apply(lambda x: [1, 2], axis = 1)

输出:

pandas-apply-4

返回类似列表的结果将是一个系列。

例 5:

# Program to illustrate the use of 
# pandas.DataFrame.apply() method

# Importing required Libraries
import pandas
import numpy

# Creating dataframe
dataFrame = pandas.DataFrame([[4, 9], ] * 3, columns =['A', 'B'])
print('Data Frame:')
display(dataFrame)

# Using pandas.DataFrame.apply() on the data frame
print('Returning multiple columns from Pandas apply()')
dataFrame.apply(lambda x: [1, 2], axis = 1, result_type ='expand')

输出:

pandas-apply-5

传递 result_type='expand '会将类似列表的结果扩展到 Dataframe 的列。

例 6:

# Program to illustrate the use of 
# pandas.DataFrame.apply() method

# Importing required Libraries
import pandas
import numpy

# Creating dataframe
dataFrame = pandas.DataFrame([[4, 9], ] * 3, columns =['A', 'B'])
print('Data Frame:')
display(dataFrame)

# Using pandas.DataFrame.apply() on the data frame
print('Returning multiple columns from Pandas apply()')
dataFrame.apply(lambda x: pandas.Series(
    [1, 2], index =['foo', 'bar']), axis = 1)

输出:

pandas-apply-6

在函数内部返回 Series 类似于传递 result_type =“expand”。产生的列名将是系列索引。

例 7:

# Program to illustrate the use of 
# pandas.DataFrame.apply() method

# Importing required Libraries
import pandas
import numpy

# Creating dataframe
dataFrame = pandas.DataFrame([[4, 9], ] * 3, columns =['A', 'B'])
print('Data Frame:')
display(dataFrame)

# Using pandas.DataFrame.apply() on the data frame
print('Returning multiple columns from Pandas apply()')
dataFrame.apply(lambda x: [1, 2], axis = 1, result_type ='broadcast')

输出:

pandas-apply-7 传递 result_type='broadcast '将保证相同的形状结果，无论是列表式还是标量由函数返回，并沿轴广播。生成的列名将是原件。

回到顶部