当前位置:网站首页>How to customize sorting for pandas dataframe

How to customize sorting for pandas dataframe

2020-11-06 01:28:09 Artificial intelligence meets pioneer

author |B. Chen compile |VK source |Towards Data Science

Pandas DataFrame There's a built-in method sort_values(), You can sort values according to a given variable . The method itself is quite simple to use , But it doesn't work with custom sort , for example ,

  • t T-shirt size :XS、S、M、L and XL

  • month : January 、 February 、 March 、 April, etc

  • What day : Monday 、 Tuesday 、 Wednesday 、 Thursday 、 Friday 、 Saturday and Sunday .

In this paper , We will learn how to deal with Pandas DataFrame Custom sort .

Please check my Github repo To get the source code :https://github.com/BindiChen/machine-learning/blob/master/data-analysis/017-pandas-custom-sort/pandas-custom-sort.ipynb

problem

Suppose we have a data set about clothing stores :

df = pd.DataFrame({
    'cloth_id': [1001, 1002, 1003, 1004, 1005, 1006],
    'size': ['S', 'XL', 'M', 'XS', 'L', 'S'],
})

We can see , Each piece of cloth has a size value , The data should be sorted in the following order :

  • XS For extra large

  • S For Trumpet

  • M For medium

  • L For big

  • XL For extra large

however , When calling sort_values('size') when , You will get the following output .

The output is not what we want , But it's technically correct . actually ,sort_values() It is to sort numerical data in numerical order , Sort the object data in alphabetical order .

Here are two common solutions :

  1. Create a new column for a custom sort

  2. Use CategoricalDtype Cast data to an ordered category type

Create a new column for a custom sort

In this solution , A mapping data frame is needed to represent a custom sort , Then create a new column from the map , Finally, we can sort the data by new columns . Let's take an example to see how this works .

First , Let's create a mapping data frame to represent a custom sort .

df_mapping = pd.DataFrame({
    'size': ['XS', 'S', 'M', 'L', 'XL'],
})

sort_mapping = df_mapping.reset_index().set_index('size')

after , Use sort_mapping Create a new column with the mapping values in size_num.

df['size_num'] = df['size'].map(sort_mapping['index'])

Last , Sort values by new column size .

df.sort_values('size_num')

This, of course, is our job . But it creates an alternate column , Efficiency may be reduced when dealing with large data sets .

We can use CategoricalDtype To solve this problem more effectively .

Use CategoricalDtype Cast data to an ordered category type

CategoricalDtype Is a type of categorical data with a category and order [1]. It's very useful for creating custom sorts [2]. Let's take an example to see how this works .

First , Let's import CategoricalDtype.

from pandas.api.types import CategoricalDtype

then , Create a custom category type cat_size_order

  • The first parameter is set to ['XS'、'S'、'M'、'L'、'XL'] As a unique value of size .

  • The second parameter ordered=True, Think of this variable as ordered .

cat_size_order = CategoricalDtype(
    ['XS', 'S', 'M', 'L', 'XL'], 
    ordered=True
)

then , call astype(cat_size_order) Cast size data to a custom category type . By running df['size'], We can see size Column has been converted to a category type , The order is [XS<S<M<L<XL].

>>> df['size'] = df['size'].astype(cat_size_order)
>>> df['size']

0     S
1    XL
2     M
3    XS
4     L
5     S
Name: size, dtype: category
Categories (5, object): [XS < S < M < L < XL]

Last , We can call the same method to sort the values .

df.sort_values('size')

It works better . Let's see what the principle is .

Use cat Of codes Attribute access

Now? size Column has been converted to category type , We can use .cat Accessor to view the classification properties . Behind the scenes , It USES codes Property to represent the size of an ordered variable .

Let's create a new column code , So we can compare size and code values side by side .

df['codes'] = df['size'].cat.codes
df

We can see XS、S、M、L and XL The codes for are 0、1、2、3、4 and 5.codes Is the actual value of the category . By running df.info(), We can see that it's actually int8.

>>> df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 6 entries, 0 to 5
Data columns (total 3 columns):
 #   Column    Non-Null Count  Dtype   
---  ------    --------------  -----   
 0   cloth_id  6 non-null      int64   
 1   size      6 non-null      category
 2   codes     6 non-null      int8    
dtypes: category(1), int64(1), int8(1)
memory usage: 388.0 bytes

Sort by multiple variables

Next , Let's make things a little more complicated . here , We will sort the data frames by multiple variables .

df = pd.DataFrame({
    'order_id': [1001, 1002, 1003, 1004, 1005, 1006, 1007],
    'customer_id': [10, 12, 12, 12, 10, 10, 10],
    'month': ['Feb', 'Jan', 'Jan', 'Feb', 'Feb', 'Jan', 'Feb'],
    'day_of_week': ['Mon', 'Wed', 'Sun', 'Tue', 'Sat', 'Mon', 'Thu'],
})

Similarly , Let's create two custom category types cat_day_of_week and cat_month, And pass them on to astype().

cat_day_of_week = CategoricalDtype(
    ['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun'], 
    ordered=True
)

cat_month = CategoricalDtype(
    ['Jan', 'Feb', 'Mar', 'Apr'], 
    ordered=True,
)

df['day_of_week'] = df['day_of_week'].astype(cat_day_of_week)
df['month'] = df['month'].astype(cat_month)

To sort by multiple variables , We just need to pass a list instead of sort_values(). for example , Press month and day_of_week Sort .

df.sort_values(['month', 'day_of_week'])

Press ustomer_id,month and day_of_week Sort .

df.sort_values(['customer_id', 'month', 'day_of_week'])

That's it , Thanks for reading .

In my, please Github Export the notebook to get the source code :https://github.com/BindiChen/machine-learning/blob/master/data-analysis/017-pandas-custom-sort/pandas-custom-sort.ipynb

Reference

Link to the original text :https://towardsdatascience.com/how-to-do-a-custom-sort-on-pandas-dataframe-ac18e7ea5320

Welcome to join us AI Blog station : http://panchuang.net/

sklearn Machine learning Chinese official documents : http://sklearn123.com/

Welcome to pay attention to pan Chuang blog resource summary station : http://docs.panchuang.net/

版权声明
本文为[Artificial intelligence meets pioneer]所创,转载请带上原文链接,感谢