一些你可能不知道的 Python 小技巧！

叁歲伎倆 2022-10-28 10:19 233阅读 0赞

> 点击上方“**机器学习与生成对抗网络**”，关注"星标"
> 
> 获取有趣、好玩的前沿干货！

作者 | Sebastian Opałczyński

编译 | 弯月   责编 | 张文

出品 | CSDN（ID：CSDNnews）

在本文中，我们来看一看日常工作中经常使用的一些 Python 小技巧。

#     #

![format_png][]

# **集合**    #

开发人员常常忘记 Python 也有集合数据类型，大家都喜欢使用列表处理一切。

集合（set）是什么？简单来说就是：**集合是一组无序事物的汇集，不包含重复元素。**

如果你熟练掌握集合及其逻辑，那么很多问题都可以迎刃而解。举个例子，如何获取一个单词中出现的字母？

myword = "NanananaBatman"
    set(myword)
    {'N', 'm', 'n', 'B', 'a', 't'}

就这么简单，问题解决了，这个例子就来自 Python 的官方文档，大可不必过于惊讶。

再举一个例子，如何获取一个列表的各个元素，且不重复？

# first you can easily change set to list and other way around
    mylist = ["a", "b", "c","c"]
    # let's make a set out of it
    myset = set(mylist)
    # myset will be:
    {'a', 'b', 'c'}
    # and, it's already iterable so you can do:
    for element in myset:
        print(element)
    # but you can also convert it to list again:
    mynewlist = list(myset)
    # and mynewlist will be:
    ['a', 'b', 'c']

我们可以看到，“c”元素不再重复出现了。只有一个地方你需要注意，mylist 与 mynewlist 之间的元素顺序可能会有所不同：

mylist = ["c", "c", "a","b"]
    mynewlist = list(set(mylist))
    # mynewlist is:
    ['a', 'b', 'c']

可以看出，两个列表的元素顺序不同。

下面，我们来进一步深入。

假设某些实体之间有一对多的关系，举个更加具体的例子：用户与权限。通常，一个用户可以拥有多个权限。现在假设某人想要修改多个权限，即同时添加和删除某些权限，应当如何解决这个问题？

# this is the set of permissions before change;
    original_permission_set = {"is_admin","can_post_entry", "can_edit_entry", "can_view_settings"}
    # this is new set of permissions;
    new_permission_set = {"can_edit_settings","is_member", "can_view_entry", "can_edit_entry"}
    # now permissions to add will be:
    new_permission_set.difference(original_permission_set)
    # which will result:
    {'can_edit_settings', 'can_view_entry', 'is_member'}
    # As you can see can_edit_entry is in both sets; so we do notneed
    # to worry about handling it
    # now permissions to remove will be:
    original_permission_set.difference(new_permission_set)
    # which will result:
    {'is_admin', 'can_view_settings', 'can_post_entry'}
    # and basically it's also true; we switched admin to member, andadd
    # more permission on settings; and removed the post_entrypermission

总的来说，不要害怕使用集合，它们能帮助你解决很多问题，更多详情，请参考 Python 官方文档。

#     #

![format_png 1][]

# **日历**    #

当开发与日期和时间有关的功能时，有些信息可能非常重要，比如某一年的这个月有多少天。这个问题看似简单，但是我相信日期和时间是一个非常有难度的话题，而且我觉得日历的实现问题非常多，简直就是噩梦，因为你需要考虑大量的极端情况。

那么，究竟如何才能找出某个月有多少天呢？

import calendar
    calendar.monthrange(2020, 12)
    # will result:
    (1, 31)
    # BUT! you need to be careful here, why? Let's read thedocumentation:
    help(calendar.monthrange)
    # Help on function monthrange in module calendar:
    # monthrange(year, month)
    #     Return weekday (0-6~ Mon-Sun) and number of days (28-31) for
    #    year, month.
    # As you can see the first value returned in tuple is a weekday,
    # not the number of the first day for a given month; let's try
    # to get the same for 2021
    calendar.monthrange(2021, 12)
    (2, 31)
    # So this basically means that the first day of December 2021 isWed
    # and the last day of December 2021 is 31 (which is obvious,cause
    # December always has 31 days)
    # let's play with February
    calendar.monthrange(2021, 2)
    (0, 28)
    calendar.monthrange(2022, 2)
    (1, 28)
    calendar.monthrange(2023, 2)
    (2, 28)
    calendar.monthrange(2024, 2)
    (3, 29)
    calendar.monthrange(2025, 2)
    (5, 28)
    # as you can see it handled nicely the leap year;

某个月的第一天当然非常简单，就是 1 号。但是，“某个月的第一天是周X”，如何使用这条信息呢？你可以很容易地查到某一天是周几：

calendar.monthrange(2024, 2)
    (3, 29)
    # means that February 2024 starts on Thursday
    # let's define simple helper:
    weekdays = ["Monday", "Tuesday","Wednesday", "Thursday", "Friday","Saturday", "Sunday"]
    # now we can do something like:
    weekdays[3]
    # will result in:
    'Thursday'
    # now simple math to tell what day is 15th of February 2020:
    offset = 3  # it's thefirst value from monthrange
    for day in range(1, 29):
        print(day,weekdays[(day + offset - 1) % 7])
    1 Thursday
    2 Friday
    3 Saturday
    4 Sunday
    ...
    18 Sunday
    19 Monday
    20 Tuesday
    21 Wednesday
    22 Thursday
    23 Friday
    24 Saturday
    ...
    28 Wednesday
    29 Thursday
    # which basically makes sense;

也许这段代码不适合直接用于生产，因为你可以使用 datetime 更容易地查找星期：

from datetime import datetime
    mydate = datetime(2024, 2, 15)
    datetime.weekday(mydate)
    # will result:
    3
    # or:
    datetime.strftime(mydate, "%A")
    'Thursday'

总的来说，日历模块有很多有意思的地方，值得慢慢学习：

# checking if year is leap:
    calendar.isleap(2021)  #False
    calendar.isleap(2024)  #True
    # or checking how many days will be leap days for given yearspan:
    calendar.leapdays(2021, 2026) # 1
    calendar.leapdays(2020, 2026) # 2
    # read the help here, as range is: [y1, y2), meaning that second
    # year is not included;
    calendar.leapdays(2020, 2024) # 1

#     #

![format_png 2][]

# **枚举有第二个参数** #

是的，枚举有第二个参数，可能很多有经验的开发人员都不知道。下面我们来看一个例子：

mylist = ['a', 'b', 'd', 'c', 'g', 'e']
    for i, item in enumerate(mylist):
        print(i, item)
    # Will give:
    0 a
    1 b
    2 d
    3 c
    4 g
    5 e
    # but, you can add a start for enumeration:
    for i, item in enumerate(mylist, 16):
        print(i, item)
    # and now you will get:
    16 a
    17 b
    18 d
    19 c
    20 g
    21 e

第二个参数可以指定枚举开始的地方，比如上述代码中的 enumerate(mylist,16)。如果你需要处理偏移量，则可以考虑这个参数。

#     #

![format_png 3][]

# **if-else 逻辑**    #

你经常需要根据不同的条件，处理不同的逻辑，经验不足的开发人员可能会编写出类似下面的代码：

OPEN = 1
    IN_PROGRESS = 2
    CLOSED = 3
    def handle_open_status():
        print('Handling openstatus')
    def handle_in_progress_status():
        print('Handling inprogress status')
    def handle_closed_status():
        print('Handling closedstatus')
    def handle_status_change(status):
        if status == OPEN:
           handle_open_status()
        elif status ==IN_PROGRESS:
            handle_in_progress_status()
        elif status == CLOSED:
           handle_closed_status()
    handle_status_change(1)  #Handling open status
    handle_status_change(2)  #Handling in progress status
    handle_status_change(3)  #Handling closed status

虽然这段代码看上去也没有那么糟，但是如果有 20 多个条件呢？

那么，究竟应该怎样处理呢？

from enum import IntEnum
    class StatusE(IntEnum):
        OPEN = 1
        IN_PROGRESS = 2
        CLOSED = 3
    def handle_open_status():
        print('Handling openstatus')
    def handle_in_progress_status():
        print('Handling inprogress status')
    def handle_closed_status():
        print('Handling closedstatus')
    handlers = {
        StatusE.OPEN.value:handle_open_status,
       StatusE.IN_PROGRESS.value: handle_in_progress_status,
        StatusE.CLOSED.value:handle_closed_status
    }
    def handle_status_change(status):
        if status not inhandlers:
             raiseException(f'No handler found for status: {status}')
        handler =handlers[status]
        handler()
    handle_status_change(StatusE.OPEN.value)  # Handling open status
    handle_status_change(StatusE.IN_PROGRESS.value)  # Handling in progress status
    handle_status_change(StatusE.CLOSED.value)  # Handling closed status
    handle_status_change(4)  #Will raise the exception

在 Python 中这种模式很常见，它可以让代码看起来更加整洁，尤其是当方法非常庞大，而且需要处理大量条件时。

#     #

![format_png 4][]

# **enum 模块**    #

enum 模块提供了一系列处理枚举的工具函数，最有意思的是 Enum 和 IntEnum。我们来看个例子：

from enum import Enum, IntEnum, Flag, IntFlag
    class MyEnum(Enum):
        FIRST ="first"
        SECOND ="second"
        THIRD ="third"
     
    class MyIntEnum(IntEnum):
        ONE = 1
        TWO = 2
        THREE = 3
    # Now we can do things like:
    MyEnum.FIRST  #<MyEnum.FIRST: 'first'>
    # it has value and name attributes, which are handy:
    MyEnum.FIRST.value  #'first'
    MyEnum.FIRST.name  #'FIRST'
    # additionally we can do things like:
    MyEnum('first')  #<MyEnum.FIRST: 'first'>, get enum by value
    MyEnum['FIRST']  #<MyEnum.FIRST: 'first'>, get enum by name

使用 IntEnum 编写的代码也差不多，但是有几个不同之处：

MyEnum.FIRST == "first"  # False
    # but
    MyIntEnum.ONE == 1  # True
    # to make first example to work:
    MyEnum.FIRST.value == "first"  # True

在中等规模的代码库中，enum 模块在管理常量方面可以提供很大的帮助。

enum 的本地化可能有点棘手，但也可以实现，我用django快速演示一下：

from enum import Enum
    from django.utils.translation import gettext_lazy as _
    class MyEnum(Enum):
        FIRST ="first"
        SECOND ="second"
        THIRD ="third"
        @classmethod
        def choices(cls):
            return [
                (cls.FIRST.value, _('first')),
                (cls.SECOND.value, _('second')),
                (cls.THIRD.value, _('third'))
             ]
    # And later in eg. model definiton:
    some_field = models.CharField(max_length=10,choices=MyEnum.choices())

![format_png 5][]

# **iPython**    #

iPython 就是交互式 Python，它是一个交互式的命令行 shell，有点像 Python 解释器。

首先，你需要安装 iPython：

pip install ipython

接下来，你只需要在输入命令的时候，将 Python 换成 ipython：

# you should see something like this after you start:
    Python 3.8.5 (default, Jul 28 2020, 12:59:40)
    Type 'copyright', 'credits' or 'license' for more information
    IPython 7.18.1 -- An enhanced Interactive Python. Type '?' forhelp.
    In [1]:

ipython 支持很多系统命令，比如 ls 或 cat，tab 键可以显示提示，而且你还可以使用上下键查找前面用过的命令。更多具体信息，请参见官方文档。

参考链接：https://levelup.gitconnected.com/python-tricks-i-can-not-live-without-87ae6aff3af8  
![format_png 6][]

**猜您喜欢：**

[**超100篇！CVPR 2020最全GAN论文梳理汇总！**][100_CVPR 2020_GAN]

[拆解组新的GAN：解耦表征MixNMatch][GAN_MixNMatch]

[StarGAN第2版：多域多样性图像生成][StarGAN_2]

[附下载 | 《可解释的机器学习》中文版][Link 1]

[附下载 |《TensorFlow 2.0 深度学习算法实战》][_TensorFlow 2.0]

[附下载 |《计算机视觉中的数学方法》分享][Link 2]

[《基于深度学习的表面缺陷检测方法综述》][Link 3]

[《零样本图像分类综述: 十年进展》][Link 4]

[《基于深度神经网络的少样本学习综述》][Link 5]

![format_png 7][]

[format_png]: /images/20221024/2b86a988b4f44c949ba36aa56a40c1d8.png
[format_png 1]: /images/20221024/c3064e0cf92246368f367ec53e519f2b.png
[format_png 2]: /images/20221024/74cfbd19c35142b8af64680b00e94f9b.png
[format_png 3]: /images/20221024/ddb9e4dad92e46dab72109c5ea295784.png
[format_png 4]: /images/20221024/b841b1104a56428da349151ffd73268a.png
[format_png 5]: /images/20221024/ce3edb7d93164e7ebc1d19f562e1626d.png
[format_png 6]: /images/20221024/bcc739cd05d84d4284fcda5dad1868ed.png
[100_CVPR 2020_GAN]: http://mp.weixin.qq.com/s?__biz=MzU5MTgzNzE0MA%3D%3D&chksm=fe29a94cc95e205ae880cb9518cb16a54c0abde69e5a8654f7fedf70a0cc990775265bf28a1c&idx=1&mid=2247486455&scene=21&sn=94156f53845cdf07e859e343862be745#wechat_redirect
[GAN_MixNMatch]: http://mp.weixin.qq.com/s?__biz=MzU5MTgzNzE0MA%3D%3D&chksm=fe29aaa9c95e23bf3131249a8d008ce3acdff7c52c7f6be84d92c6d607ca804b5ceb2a4222e3&idx=1&mid=2247485458&scene=21&sn=9f562c1de84dada67268f88c72ff77ff#wechat_redirect
[StarGAN_2]: http://mp.weixin.qq.com/s?__biz=MzU5MTgzNzE0MA%3D%3D&chksm=fe29ab6ec95e22785abb2afbb2e2b1c95780a4d6bfceeb915118440b41a81769a8776dcc62b7&idx=1&mid=2247485909&scene=21&sn=048d1336b8083bac3755e27a1f968d06#wechat_redirect
[Link 1]: http://mp.weixin.qq.com/s?__biz=MzU5MTgzNzE0MA%3D%3D&chksm=fe29a4bac95e2dac758c4aac10a3aac49eb062bc0829d4ef2ebc8e6d5069d246385c5084a2b4&idx=1&mid=2247484929&scene=21&sn=523220858287ed3e6105919703c8e833#wechat_redirect
[_TensorFlow 2.0]: http://mp.weixin.qq.com/s?__biz=MzU5MTgzNzE0MA%3D%3D&chksm=fe29a5b0c95e2ca6599d4a91d5d40900566b7525de3dd85259e7a6c173aa26aa7ea3641d1aaf&idx=2&mid=2247485195&scene=21&sn=e3387067db29d1759755c6c2ea96c870#wechat_redirect
[Link 2]: http://mp.weixin.qq.com/s?__biz=MzU5MTgzNzE0MA%3D%3D&chksm=fe29aa8fc95e23993e057fb0dbe09f2810ea07184a439c173f404a7ab657cadcde3d24efa52a&idx=2&mid=2247485492&scene=21&sn=c4fb135de02cd52f06bc61b0e79476e4#wechat_redirect
[Link 3]: http://mp.weixin.qq.com/s?__biz=MzU5MTgzNzE0MA%3D%3D&chksm=fe29aa8fc95e2399a93dae3af3b13626643b18cf42842d1a7ada0de46486fa8b865aa2421d2f&idx=1&mid=2247485492&scene=21&sn=6f8af433521b5fee5aceab62e0111115#wechat_redirect
[Link 4]: http://mp.weixin.qq.com/s?__biz=MzU5MTgzNzE0MA%3D%3D&chksm=fe29a44fc95e2d597fc88ca7f10cae079f821deedb6f93b10f0985aa8f6f7c06e369d716f785&idx=1&mid=2247485172&scene=21&sn=f65b58ed3b4c1d09163271069dce4738#wechat_redirect
[Link 5]: http://mp.weixin.qq.com/s?__biz=MzU5MTgzNzE0MA%3D%3D&chksm=fe29a7e6c95e2ef0a6649c0c904fa28b61b4ac76774df4acdce8fff3a909369c7c2f8d22a4b5&idx=1&mid=2247484765&scene=21&sn=b1b9a9ef6c923abad6cf38364105264c#wechat_redirect
[format_png 7]: /images/20221024/e0090c18b8b54356801b23de55808686.png