周萝卜

V1

2022/08/23阅读:19主题:山吹

Pandas 重置索引深度总结

今天我们来讨论 Pandas 中的 reset_index() 方法,包括为什么我们需要在 Pandas 中重置 DataFrame 的索引,以及我们应该如何应用该方法

在本文我们将使用 Kaggle 上的数据集样本 Animal Shelter Analytics 来作为我们的测试数据

Pandas 中的 Reset_Index() 是什么?

如果我们使用 Pandas 的 read_csv() 方法读取 csv 文件而不指定任何索引,则生成的 DataFrame 将具有默认的基于整数的索引,第一行从 0 开始,随后每行增加 1:

import pandas as pd
import numpy as np

df = pd.read_csv('Austin_Animal_Center_Intakes.csv').head()
df

Output:

	Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
3	A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
4	A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

在某些情况下,我们可能希望拥有更有意义的行标签,因此我们将选择 DataFrame 的其中一列作为 DataFrame 索引。我们可以使用 read_csv() 方法的 index_col 参数直接执行此操作:

df = pd.read_csv('Austin_Animal_Center_Intakes.csv', index_col='Animal ID').head()
df

Output:

Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

或者我们还可以使用 set_index() 方法将 DataFrame 的任何列设置为 DataFrame 索引:

df = pd.read_csv('Austin_Animal_Center_Intakes.csv').head()
df.set_index('Animal ID', inplace=True)
df

Output:

Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

如果在某个时候我们需要恢复默认的数字索引呢,这时就可以使用 reset_index()函数了

df.reset_index()

Output:

	Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
3	A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
4	A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

此方法的默认行为包括用默认的基于整数的索引替换现有的 DataFrame 索引,并将旧索引转换为与旧索引同名的新列(或名称索引)。此外,默认情况下,reset_index() 方法会从 MultiIndex 中删除所有级别并且不会影响原始 DataFrame 数据,而是创建一个新的

何时使用 Reset_Index() 方法

reset_index() 方法将 DataFrame 索引重置为默认数字索引,在以下情况下特别有用:

  • 执行数据整理时——尤其是过滤数据或删除缺失值等预处理操作,会导致较小的 DataFrame 具有不再连续的数字索引
  • 当索引应该被视为一个常见的 DataFrame 列时
  • 当索引标签没有提供有关数据的任何有价值的信息时

如何调整 Reset_Index() 方法

前面的讨论中,我们看到了当我们不向它传递任何参数时,reset_index() 方法是如何工作的,当然如果有需要,我们可以通过调整方法的各种参数来更改此默认行为。 让我们看看最有用的三种参数:level、drop 和 inplace

level

此参数采用整数、字符串、元组或列表作为可能的数据类型,并且仅适用于具有 MultiIndex 的 DataFrame,如下所示:

df_multiindex = pd.read_csv('Austin_Animal_Center_Intakes.csv', index_col=['Animal ID''Name']).head()
df_multiindex

Output:

Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

事实上,如果现在我们检查上面 DataFrame 的索引,我们会发现它不是一个常见的 DataFrame 索引,而是一个 MultiIndex 对象:

df_multiindex.index

Output:

MultiIndex([('A786884',  '*Brock'),
            ('A706918',   'Belle'),
            ('A724273', 'Runster'),
            ('A665644',       nan),
            ('A682524',     'Rio')],
           names=['Animal ID', 'Name'])

默认情况下,reset_index() 方法参数 level (level=None) 会移除 MultiIndex 的所有级别:

df_multiindex.reset_index()

Output:

	Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
3	A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
4	A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

我们看到 DataFrame 的两个索引都被转换为通用 DataFrame 列,而索引被重置为默认的基于整数的索引

相反,如果我们显式传递 level 的值,则此参数会从 DataFrame 索引中删除选定的级别,并将它们作为常见的 DataFrame 列返回(除非我们选择使用 drop 参数从 DataFrame 中完全删除此信息)。比较以下操作:

df_multiindex.reset_index(level='Animal ID')

Output:

Name	Animal ID	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
*Brock	A786884	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
Belle	A706918	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
Runster	A724273	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
NaN	A665644	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
Rio	A682524	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

最开始 Animal ID 是 DataFrame 的索引之一,当我们设置 level 参数后,将其从索引中删除并作为称为 Animal ID 的公共列插入到 DataFrame 中

df_multiindex.reset_index(level='Name')

Output:

Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

在这里,Name 最初是 DataFrame 的索引之一,设置完level参数后,就变成了一个常用的列,叫做Name

drop

此参数决定在索引重置后是否将旧索引保留为通用 DataFrame 列,或者将其从 DataFrame 中完全删除。默认情况下 (drop=False) 是进行保留的,正如我们在前面的所有示例中看到的那样。否则,如果我们不想将旧索引保留为列,我们可以在索引重置后将其从 DataFrame 中完全删除(drop=True):

df

Output:

Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

增加 drop 参数

df.reset_index(drop=True)

Output:

	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
3	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
4	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

在上面的 DataFrame 中,旧索引中包含的信息已完全从 DataFrame 中删除了

drop 参数也适用于具有 MultiIndex 的 DataFrame,就像我们之前创建的那样:

df_multiindex

Output:

Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

增加 drop 参数

df_multiindex.reset_index(drop=True)

Output:

	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
3	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
4	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

两个旧索引都已从 Dataframe 中完全删除,并且索引已重置为默认值

当然,我们可以结合 drop 和 level 参数,指定要从 DataFrame 中完全删除哪些旧索引:

df_multiindex.reset_index(level='Animal ID', drop=True)

Output:

	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
Name										
*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

旧索引 Animal ID 已从索引和 DataFrame 本身中删除,另一个索引 Name 被保留为 DataFrame 的当前索引

inplace

该参数决定是直接修改原来的 DataFrame 还是新建一个 DataFrame 对象。默认情况下,它会使用新索引 (inplace=False) 创建一个新的 DataFrame,并保持原始 DataFrame 不变。让我们使用默认参数再次运行 reset_index() 方法,然后将结果与原始 DataFrame 进行比较:

df.reset_index()

Output:

	Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
3	A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
4	A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray
df

Output:

Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

即使我们将索引重置为运行第一段代码的默认数字,原始 DataFrame 仍然保持不变。 如果我们需要将原始 DataFrame 重新分配给对其应用 reset_index() 方法的结果,我们可以直接重新分配它(df = df.reset_index())或将参数 inplace=True 传递给该方法:

df.reset_index(inplace=True)
df

Output:

	Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
3	A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
4	A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

我们看到现在更改已直接应用于原始 DataFrame 之上了

应用实例:删除缺失值后重置索引

让我们将到目前为止讨论的所有内容付诸实践,看看当我们从 DataFrame 中删除缺失值时,重置 DataFrame 索引是如何有用的

首先,让我们恢复我们最开始时创建的第一个 DataFrame,它具有默认数字索引:

df = pd.read_csv('Austin_Animal_Center_Intakes.csv').head()
df

Output:

	Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
3	A665644	NaN	10/21/2013 07:59:00 AM	10/21/2013 07:59:00 AM	Austin (TX)	Stray	Sick	Cat	Intact Female	4 weeks	Domestic Shorthair Mix	Calico
4	A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

我们看到 DataFrame 中有一个缺失值,让我们使用 dropna() 方法删除具有缺失值的整行

df.dropna(inplace=True)
df

Output:

	Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
4	A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

该行已从 DataFrame 中删除,但是索引不再是连续的:0、1、2、4。让我们重新设置它:

df.reset_index()

Output:

	index	Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	0	A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	1	A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	2	A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
3	4	A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

现在索引是连续的了,但是由于我们没有显式传递 drop 参数,旧索引被转换为列,具有默认名称 index,下面让我们从 DataFrame 中完全删除旧索引:

df.reset_index(drop=True)

Output:

	Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
3	A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

现在我们彻底摆脱了无意义的旧索引,当前索引是连续的。最后一步是使用 inplace 参数将这些修改保存到我们的原始 DataFrame 中:

df.reset_index(drop=True, inplace=True)
df

Output:

	Animal ID	Name	DateTime	MonthYear	Found Location	Intake Type	Intake Condition	Animal Type	Sex upon Intake	Age upon Intake	Breed	Color
0	A786884	*Brock	01/03/2019 04:19:00 PM	01/03/2019 04:19:00 PM	2501 Magin Meadow Dr in Austin (TX)	Stray	Normal	Dog	Neutered Male	2 years	Beagle Mix	Tricolor
1	A706918	Belle	07/05/2015 12:59:00 PM	07/05/2015 12:59:00 PM	9409 Bluegrass Dr in Austin (TX)	Stray	Normal	Dog	Spayed Female	8 years	English Springer Spaniel	White/Liver
2	A724273	Runster	04/14/2016 06:43:00 PM	04/14/2016 06:43:00 PM	2818 Palomino Trail in Austin (TX)	Stray	Normal	Dog	Intact Male	11 months	Basenji Mix	Sable/White
3	A682524	Rio	06/29/2014 10:38:00 AM	06/29/2014 10:38:00 AM	800 Grove Blvd in Austin (TX)	Stray	Normal	Dog	Neutered Male	4 years	Doberman Pinsch/Australian Cattle Dog	Tan/Gray

总结

今天我们从多个方面讨论了 reset_index() 方法

  • reset_index() 方法的默认行为
  • 如何恢复 DataFrame 的默认数字索引
  • 何时使用 reset_index() 方法
  • 该方法最重要的几个参数
  • 如何使用 MultiIndex
  • 如何从 DataFrame 中完全删除旧索引
  • 如何将修改直接保存到原始 DataFrame 中

最好我们又完整的完成了一个在删除缺失值后重置 DataFrame 索引的实战案例

好了,这就是今天分享的全部内容,喜欢就点个赞吧

分类:

后端

标签:

Python

作者介绍

周萝卜
V1