En AZ med nyttige Python-tricks

Python er et af verdens mest populære programmeringssprog efterspurgt. Dette er af mange grunde:

  • det er let at lære
  • det er super alsidigt
  • den har et stort udvalg af moduler og biblioteker

Jeg bruger Python dagligt som en integreret del af mit job som dataforsker. Undervejs har jeg hentet et par nyttige tricks og tip.

Her har jeg delt nogle af dem i et AZ-format.

De fleste af disse 'tricks' er ting, jeg har brugt eller snublet over under mit daglige arbejde. Nogle fandt jeg, mens jeg gennemsøgte Python Standard Library-dokumenterne. Et par andre fandt jeg ved at søge gennem PyPi.

Dog kredit, hvor det skyldes - Jeg opdagede fire eller fem af dem på awesome-python.com. Dette er en kurateret liste over hundredvis af interessante Python-værktøjer og moduler. Det er værd at søge efter inspiration!

alle eller nogen

En af de mange grunde til, at Python er et så populært sprog, er fordi det er læsbart og udtryksfuldt.

Det jacks ofte, at Python er 'eksekverbar pseudokode'. Men når du kan skrive kode som denne, er det svært at argumentere ellers:

x = [True, True, False] if any(x): print("At least one True") if all(x): print("Not one False") if any(x) and not all(x): print("At least one True and one False")

bashplotlib

Vil du plotte grafer i konsollen?

$ pip install bashplotlib

Du kan have grafer i konsollen.

samlinger

Python har nogle gode standarddatatyper, men nogle gange opfører de sig ikke nøjagtigt, som du vil have dem til.

Heldigvis tilbyder Python Standard Library samlingsmodulet. Denne praktiske tilføjelse giver dig yderligere datatyper.

from collections import OrderedDict, Counter # Remembers the order the keys are added! x = OrderedDict(a=1, b=2, c=3) # Counts the frequency of each character y = Counter("Hello World!") 

dir

Har du nogensinde spekuleret på, hvordan du kan se inde i et Python-objekt og se, hvilke egenskaber det har? Selvfølgelig har du det.

Fra kommandolinjen:

>>> dir() >>> dir("Hello World") >>> dir(dir)

Dette kan være en virkelig nyttig funktion, når du kører Python interaktivt og til dynamisk udforskning af objekter og moduler, du arbejder med.

Læs mere her.

emoji

Ja virkelig.

$ pip install emoji

Lad ikke som om du ikke prøver det ...

from emoji import emojize print(emojize(":thumbs_up:"))

?

fra __future__ import

En konsekvens af Pythons popularitet er, at der altid er nye versioner under udvikling. Nye versioner betyder nye funktioner - medmindre din version er forældet.

Frygt dog ikke. Modulet __future__ giver dig mulighed for at importere funktionalitet fra fremtidige versioner af Python. Det er bogstaveligt talt som tidsrejser eller magi eller noget.

from __future__ import print_function print("Hello World!")

Hvorfor ikke prøve at importere krøllede seler?

geopy

Geografi kan være et udfordrende terræn for programmører at navigere (ha, en ordspil!). Men geopimodulet gør det unødvendigt nemt.

$ pip install geopy

Det fungerer ved at abstrahere API'erne for en række forskellige geokodningstjenester. Det giver dig mulighed for at få et sted med fuld gadenavn, bredde, længdegrad og endda højde.

Der er også en nyttig afstandsklasse. Den beregner afstanden mellem to placeringer i din foretrukne måleenhed.

from geopy import GoogleV3 place = "221b Baker Street, London" location = GoogleV3().geocode(place) print(location.address) print(location.location)

hvordan gør jeg

Sidder du ved et kodningsproblem og kan du ikke huske den løsning, du så før? Har du brug for at kontrollere StackOverflow, men ikke ønsker at forlade terminalen?

Så har du brug for dette nyttige kommandolinjeværktøj.

$ pip install howdoi

Stil det uanset spørgsmål, du har, og det vil gøre sit bedste for at returnere et svar.

$ howdoi vertical align css $ howdoi for loop in java $ howdoi undo commits in git

Vær dog opmærksom - det skraber kode fra de bedste svar fra StackOverflow. Det giver måske ikke altid de mest nyttige oplysninger ...

$ howdoi exit vim

inspicere

Pythons inspektionsmodul er fantastisk til at forstå, hvad der sker bag kulisserne. Du kan endda kalde dens metoder på sig selv!

Kodeprøven nedenfor bruger inspect.getsource()til at udskrive sin egen kildekode. Det bruger også inspect.getmodule()til at udskrive det modul, hvor det blev defineret.

Den sidste linie kode udskriver sit eget linjenummer.

import inspect print(inspect.getsource(inspect.getsource)) print(inspect.getmodule(inspect.getmodule)) print(inspect.currentframe().f_lineno)

Of course, beyond these trivial uses, the inspect module can prove useful for understanding what your code is doing. You could also use it for writing self-documenting code.

Jedi

The Jedi library is an autocompletion and code analysis library. It makes writing code quicker and more productive.

Unless you’re developing your own IDE, you’ll probably be most interested in using Jedi as an editor plugin. Luckily, there are already loads available!

You may already be using Jedi, however. The IPython project makes use of Jedi for its code autocompletion functionality.

**kwargs

When learning any language, there are many milestones along the way. With Python, understanding the mysterious **kwargs syntax probably counts as one.

The double-asterisk in front of a dictionary object lets you pass the contents of that dictionary as named arguments to a function.

The dictionary’s keys are the argument names, and the values are the values passed to the function. You don’t even need to call it kwargs!

dictionary = {"a": 1, "b": 2} def someFunction(a, b): print(a + b) return # these do the same thing: someFunction(**dictionary) someFunction(a=1, b=2)

This is useful when you want to write functions that can handle named arguments not defined in advance.

List comprehensions

One of my favourite things about programming in Python are its list comprehensions.

These expressions make it easy to write very clean code that reads almost like natural language.

You can read more about how to use them here.

numbers = [1,2,3,4,5,6,7] evens = [x for x in numbers if x % 2 is 0] odds = [y for y in numbers if y not in evens] cities = ['London', 'Dublin', 'Oslo'] def visit(city): print("Welcome to "+city) for city in cities: visit(city)

map

Python supports functional programming through a number of inbuilt features. One of the most useful is the map() function — especially in combination with lambda functions.

x = [1, 2, 3] y = map(lambda x : x + 1 , x) # prints out [2,3,4]print(list(y))

In the example above, map() applies a simple lambda function to each element in x. It returns a map object, which can be converted to some iterable object such as a list or tuple.

newspaper3k

If you haven’t seen it already, then be prepared to have your mind blown by Python’s newspaper module.

It lets you retrieve news articles and associated meta-data from a range of leading international publications. You can retrieve images, text and author names.

It even has some inbuilt NLP functionality.

So if you were thinking of using BeautifulSoup or some other DIY webscraping library for your next project, save yourself the time and effort and $ pip install newspaper3k instead.

Operator overloading

Python provides support for operator overloading, which is one of those terms that make you sound like a legit computer scientist.

It’s actually a simple concept. Ever wondered why Python lets you use the + operator to add numbers and also to concatenate strings? That’s operator overloading in action.

You can define objects which use Python’s standard operator symbols in their own specific way. This lets you use them in contexts relevant to the objects you’re working with.

class Thing: def __init__(self, value): self.__value = value def __gt__(self, other): return self.__value > other.__value def __lt__(self, other): return self.__value  nothing # False something < nothing # Error something + nothing

pprint

Python’s default print function does its job. But try printing out any large, nested object, and the result is rather ugly.

Here’s where the Standard Library’s pretty-print module steps in. This prints out complex structured objects in an easy-to-read format.

A must-have for any Python developer who works with non-trivial data structures.

import requests import pprint url = '//randomuser.me/api/?results=1' users = requests.get(url).json() pprint.pprint(users)

Queue

Python supports multithreading, and this is facilitated by the Standard Library’s Queue module.

This module lets you implement queue data structures. These are data structures that let you add and retrieve entries according to a specific rule.

‘First in, first out’ (or FIFO) queues let you retrieve objects in the order they were added. ‘Last in, first out’ (LIFO) queues let you access the most recently added objects first.

Finally, priority queues let you retrieve objects according to the order in which they are sorted.

Here’s an example of how to use queues for multithreaded programming in Python.

__repr__

When defining a class or an object in Python, it is useful to provide an ‘official’ way of representing that object as a string. For example:

>>> file = open('file.txt', 'r') >>> print(file) 

This makes debugging code a lot easier. Add it to your class definitions as below:

class someClass: def __repr__(self): return "" someInstance = someClass() # prints  print(someInstance)

sh

Python makes a great scripting language. Sometimes using the standard os and subprocess libraries can be a bit of a headache.

The sh library provides a neat alternative.

It lets you call any program as if it were an ordinary function — useful for automating workflows and tasks, all from within Python.

import sh sh.pwd() sh.mkdir('new_folder') sh.touch('new_file.txt') sh.whoami() sh.echo('This is great!')

Type hints

Python is a dynamically-typed language. You don’t need to specify datatypes when you define variables, functions, classes etc.

This allows for rapid development times. However, there are few things more annoying than a runtime error caused by a simple typing issue.

Since Python 3.5, you have the option to provide type hints when defining functions.

def addTwo(x : Int) -> Int: return x + 2

You can also define type aliases:

from typing import List
Vector = List[float]Matrix = List[Vector]
def addMatrix(a : Matrix, b : Matrix) -> Matrix: result = [] for i,row in enumerate(a): result_row =[] for j, col in enumerate(row): result_row += [a[i][j] + b[i][j]] result += [result_row] return result x = [[1.0, 0.0], [0.0, 1.0]] y = [[2.0, 1.0], [0.0, -2.0]] z = addMatrix(x, y)

Although not compulsory, type annotations can make your code easier to understand.

They also allow you to use type checking tools to catch those stray TypeErrors before runtime. Probably worthwhile if you are working on large, complex projects!

uuid

A quick and easy way to generate Universally Unique IDs (or ‘UUIDs’) is through the Python Standard Library’s uuid module.

import uuid user_id = uuid.uuid4() print(user_id)

This creates a randomized 128-bit number that will almost certainly be unique.

In fact, there are over 2¹²² possible UUIDs that can be generated. That’s over five undecillion (or 5,000,000,000,000,000,000,000,000,000,000,000,000).

The probability of finding duplicates in a given set is extremely low. Even with a trillion UUIDs, the probability of a duplicate existing is much, much less than one-in-a-billion.

Pretty good for two lines of code.

Virtual environments

This is probably my favorite Python thing of all.

Chances are you are working on multiple Python projects at any one time. Unfortunately, sometimes two projects will rely on different versions of the same dependency. Which do you install on your system?

Luckily, Python’s support for virtual environments lets you have the best of both worlds. From the command line:

python -m venv my-project source my-project/bin/activate pip install all-the-modules 

Now you can have standalone versions and installations of Python running on the same machine. Sorted!

wikipedia

Wikipedia has a great API that allows users programmatic access to an unrivalled body of completely free knowledge and information.

The wikipedia module makes accessing this API almost embarrassingly convenient.

import wikipedia result = wikipedia.page('freeCodeCamp') print(result.summary) for link in result.links: print(link)

Like the real site, the module provides support for multiple languages, page disambiguation, random page retrieval, and even has a donate() method.

xkcd

Humour is a key feature of the Python language — after all, it is named after the British comedy sketch show Monty Python’s Flying Circus. Much of Python’s official documentation references the show’s most famous sketches.

The sense of humour isn’t restricted to the docs, though. Have a go running the line below:

import antigravity

Never change, Python. Never change.

YAML

YAML stands for ‘YAML Ain’t Markup Language’. It is a data formatting language, and is a superset of JSON.

Unlike JSON, it can store more complex objects and refer to its own elements. You can also write comments, making it particularly suited to writing configuration files.

The PyYAML module lets you use YAML with Python. Install with:

$ pip install pyyaml

And then import into your projects:

import yaml

PyYAML lets you store Python objects of any datatype, and instances of any user-defined classes also.

zip

Et sidste trick for dig, og det er virkelig sejt. Har du nogensinde haft brug for at danne en ordbog ud af to lister?

keys = ['a', 'b', 'c'] vals = [1, 2, 3] zipped = dict(zip(keys, vals))

Den zip()indbyggede funktion tager et antal gentagelige objekter og returnerer en liste over tupler. Hver tuple grupperer elementerne i inputobjekterne efter deres positionsindeks.

Du kan også 'pakke ud' objekter ved at kalde *zip()på dem.

Tak for læsningen!

Så der har du det, en AZ med Python-tricks - forhåbentlig har du fundet noget nyttigt til dit næste projekt.

Python er et meget forskelligt og veludviklet sprog, så der er sikkert mange funktioner, som jeg ikke har fået med til at inkludere.

Del venligst et af dine egne yndlings Python-tricks ved at efterlade et svar nedenfor!