Comprehensive Python Cheatsheet

^{Download text file, Fork me on GitHub or Check out FAQ.}

Main

if __name__ == '__main__':
    main()

List

<list> = <list>[from_inclusive : to_exclusive : ±step_size]

<list>.append(<el>)            # Or: <list> += [<el>]
<list>.extend(<collection>)    # Or: <list> += <collection>

<list>.sort()
<list>.reverse()
<list> = sorted(<collection>)
<iter> = reversed(<list>)

sum_of_elements  = sum(<collection>)
elementwise_sum  = [sum(pair) for pair in zip(list_a, list_b)]
sorted_by_second = sorted(<collection>, key=lambda el: el[1])
sorted_by_both   = sorted(<collection>, key=lambda el: (el[1], el[0]))
flatter_list     = list(itertools.chain.from_iterable(<list>))
product_of_elems = functools.reduce(lambda out, x: out * x, <collection>)
list_of_chars    = list(<str>)

index = <list>.index(<el>)     # Returns first index of item.
<list>.insert(index, <el>)     # Inserts item at index and moves the rest to the right.
<el> = <list>.pop([index])     # Removes and returns item at index or from the end.
<list>.remove(<el>)            # Removes first occurrence of item or raises ValueError.
<list>.clear()                 # Removes all items.

Dictionary

<view> = <dict>.keys()
<view> = <dict>.values()
<view> = <dict>.items()

value  = <dict>.get(key, default=None)          # Returns default if key does not exist.
value  = <dict>.setdefault(key, default=None)   # Same, but also adds default to dict.
<dict> = collections.defaultdict(<type>)        # Creates a dict with default value of type.
<dict> = collections.defaultdict(lambda: 1)     # Creates a dict with default value 1.

<dict>.update(<dict>)                           # Or: dict_a = {**dict_a, **dict_b}.
<dict> = dict(<collection>)                     # Creates a dict from coll. of key-value pairs.
<dict> = dict(zip(keys, values))                # Creates a dict from two collections.
<dict> = dict.fromkeys(keys [, value])          # Creates a dict from collection of keys.

value = <dict>.pop(key)                         # Removes item from dictionary.
{k: v for k, v in <dict>.items() if k in keys}  # Filters dictionary by keys.

Counter

>>> from collections import Counter
>>> colors = ['red', 'blue', 'yellow', 'blue', 'red', 'blue']
>>> counter = Counter(colors)
Counter({'blue': 3, 'red': 2, 'yellow': 1})
>>> counter.most_common()[0]
('blue', 3)

Set

<set> = set()

<set>.add(<el>)                               # Or: <set> |= {<el>}
<set>.update(<collection>)                    # Or: <set> |= <set>

<set>  = <set>.union(<coll.>)                 # Or: <set> | <set>
<set>  = <set>.intersection(<coll.>)          # Or: <set> & <set>
<set>  = <set>.difference(<coll.>)            # Or: <set> - <set>
<set>  = <set>.symmetric_difference(<coll.>)  # Or: <set> ^ <set>
<bool> = <set>.issubset(<coll.>)              # Or: <set> <= <set>
<bool> = <set>.issuperset(<coll.>)            # Or: <set> >= <set>

<set>.remove(<el>)   # Throws error.
<set>.discard(<el>)  # Doesn't throw error.

Frozenset

Is hashable, meaning it can be used as a key in a dictionary or as an element in a set.

<frozenset> = frozenset(<collection>)

Range

<range> = range(to_exclusive)
<range> = range(from_inclusive, to_exclusive)
<range> = range(from_inclusive, to_exclusive, ±step_size)

from_inclusive = <range>.start
to_exclusive   = <range>.stop

Enumerate

for i, el in enumerate(<collection> [, i_start]):
    ...

Named Tuple

Tuple is an immutable and hashable list.
Named tuple is its subclass with named elements.

>>> from collections import namedtuple
>>> Point = namedtuple('Point', 'x y')
>>> p = Point(1, y=2)
Point(x=1, y=2)
>>> p[0]
1
>>> p.x
1
>>> getattr(p, 'y')
2
>>> p._fields  # Or: Point._fields
('x', 'y')

Iterator

In this cheatsheet '<collection>' can also mean an iterator.

from itertools import count, repeat, cycle, chain, islice

<iter> = iter(<collection>)
<iter> = iter(<function>, to_exclusive)     # Sequence of return values until 'to_exclusive'.
<el>   = next(<iter> [, default])           # Raises StopIteration or returns 'default' on end.

<iter> = count(start=0, step=1)             # Returns incremented value endlessly.
<iter> = repeat(<el> [, times])             # Returns element endlessly or 'times' times.
<iter> = cycle(<collection>)                # Repeats the sequence indefinitely.

<iter> = chain(<coll.>, <coll.>, ...)       # Empties collections in order.
<iter> = chain.from_iterable(<collection>)  # Empties collections inside a collection in order.

<iter> = islice(<collection>, to_exclusive)
<iter> = islice(<collection>, from_inclusive, to_exclusive)
<iter> = islice(<collection>, from_inclusive, to_exclusive, step_size)

Generator

Convenient way to implement the iterator protocol.

def count(start, step):
    while True:
        yield start
        start += step

>>> counter = count(10, 2)
>>> next(counter), next(counter), next(counter)
(10, 12, 14)

Type

<type> = type(<el>)  # <class 'int'> / <class 'str'> / ...

from numbers import Integral, Rational, Real, Complex, Number
<bool> = isinstance(<el>, Number)

<bool> = callable(<el>)

String

<str>  = <str>.strip()                       # Strips all whitespace characters from both ends.
<str>  = <str>.strip('<chars>')              # Strips all passed characters from both ends.

<list> = <str>.split()                       # Splits on any whitespace character.
<list> = <str>.split(sep=None, maxsplit=-1)  # Splits on 'sep' str at most 'maxsplit' times.
<str>  = <str>.join(<collection>)            # Joins elements using string as separator.

<str>  = <str>.replace(old, new [, count])   # Replaces 'old' with 'new' at most 'count' times.
<bool> = <str>.startswith(<sub_str>)         # Pass tuple of strings for multiple options.
<bool> = <str>.endswith(<sub_str>)           # Pass tuple of strings for multiple options.
<int>  = <str>.index(<sub_str>)              # Returns start index of first match.

<bool> = <str>.isnumeric()                   # True if str contains only numeric characters.
<list> = textwrap.wrap(<str>, width)         # Nicely breaks string into lines.

Char

<str> = chr(<int>)  # Converts int to unicode char.
<int> = ord(<str>)  # Converts unicode char to int.

>>> ord('0'), ord('9')
(48, 57)
>>> ord('A'), ord('Z')
(65, 90)
>>> ord('a'), ord('z')
(97, 122)

Regex

import re
<str>   = re.sub(<regex>, new, text, count=0)  # Substitutes all occurrences.
<list>  = re.findall(<regex>, text)            # Returns all occurrences.
<list>  = re.split(<regex>, text, maxsplit=0)  # Use brackets in regex to keep the matches.
<Match> = re.search(<regex>, text)             # Searches for first occurrence of pattern.
<Match> = re.match(<regex>, text)              # Searches only at the beginning of the text.
<iter>  = re.finditer(<regex>, text)           # Returns all occurrences as match objects.

Parameter 'flags=re.IGNORECASE' can be used with all functions.
Parameter 'flags=re.DOTALL' makes dot also accept newline.
Use r'\1' or '\\1' for backreference.
Use '?' to make an operator non-greedy.

Match Object

<str>   = <Match>.group()   # Whole match.
<str>   = <Match>.group(1)  # Part in first bracket.
<tuple> = <Match>.groups()  # All bracketed parts.
<int>   = <Match>.start()   # Start index of a match.
<int>   = <Match>.end()     # Exclusive end index of a match.

Special Sequences

Expressions below hold true for strings that contain only ASCII characters. Use capital letters for negation.

'\d' == '[0-9]'          # Digit
'\s' == '[ \t\n\r\f\v]'  # Whitespace
'\w' == '[a-zA-Z0-9_]'   # Alphanumeric

Format

<str> = f'{<el_1>}, {<el_2>}'
<str> = '{}, {}'.format(<el_1>, <el_2>)

>>> Person = collections.namedtuple('Person', 'name height')
>>> person = Person('Jean-Luc', 187)
>>> f'{person.height}'
'187'
>>> '{p.height}'.format(p=person)
'187'

General Options

{<el>:<10}       # '<el>      '
{<el>:>10}       # '      <el>'
{<el>:^10}       # '   <el>   '
{<el>:->10}      # '------<el>'
{<el>:>0}        # '<el>'

String Options

'!r' calls object's repr() method, instead of format(), to get a string.

{'abcde'!r:<10}  # "'abcde'   "
{'abcde':.3}     # 'abc'
{'abcde':10.3}   # 'abc       '

Float Options

{1.23456:.3f}    # '1.235'
{1.23456:10.3f}  # '     1.235'
{1.23456:10.3e}  # ' 1.235e+00'
{1.23456:10.3%}  # '  123.456%'

Int Options

{ 123456:10,}    # '   123,456'
{ 123456:10_}    # '   123_456'
{ 123456:+10}    # '   +123456'
{-123456:=10}    # '-   123456'
{ 123456: }      # ' 123456'
{-123456: }      # '-123456'

{90:c}           # 'Z'
{90:X}           # '5A'
{3:08b}          # '00000011'

Numbers

Basic Functions

<num>  = pow(<num>, <num>)  # Or: <num> ** <num>
<real> = abs(<num>)
<int>  = round(<real>)
<real> = round(<real>, ±ndigits)

Math

from math import e, pi
from math import cos, acos, sin, asin, tan, atan, degrees, radians
from math import log, log10, log2
from math import inf, nan, isinf, isnan

Statistics

from statistics import mean, median, variance, pvariance, pstdev

Random

from random import random, randint, choice, shuffle
<float> = random()
<int>   = randint(from_inclusive, to_inclusive)
<el>    = choice(<list>)
shuffle(<list>)

Combinatorics

Every function returns an iterator.
If you want to print the iterator, you need to pass it to the list() function!

from itertools import product, combinations, combinations_with_replacement, permutations

>>> product([0, 1], repeat=3)
[(0, 0, 0), (0, 0, 1), (0, 1, 0), (0, 1, 1),
 (1, 0, 0), (1, 0, 1), (1, 1, 0), (1, 1, 1)]

>>> product('ab', '12')
[('a', '1'), ('a', '2'),
 ('b', '1'), ('b', '2')]

>>> combinations('abc', 2)
[('a', 'b'), ('a', 'c'), ('b', 'c')]

>>> combinations_with_replacement('abc', 2)
[('a', 'a'), ('a', 'b'), ('a', 'c'),
 ('b', 'b'), ('b', 'c'),
 ('c', 'c')]

>>> permutations('abc', 2)
[('a', 'b'), ('a', 'c'),
 ('b', 'a'), ('b', 'c'),
 ('c', 'a'), ('c', 'b')]

Datetime

Module 'datetime' provides 'date' <D>, 'time' <T>, 'datetime' <DT> and 'timedelta' <TD> classes. All are immutable and hashable.
Time and datetime can be 'aware' <a>, meaning they have defined timezone, or 'naive' <n>, meaning they don't.
If object is naive it is presumed to be in system's timezone.

from datetime import date, time, datetime, timedelta
from dateutil.tz import UTC, tzlocal, gettz

Constructors

<D>  = date(year, month, day)
<T>  = time(hour=0, minute=0, second=0, microsecond=0, tzinfo=None, fold=0)
<DT> = datetime(year, month, day, hour=0, minute=0, second=0, ...)
<TD> = timedelta(days=0, seconds=0, microseconds=0, milliseconds=0, 
                 minutes=0, hours=0, weeks=0)

Use '<D/DT>.weekday()' to get the day of the week (Mon == 0).
'fold=1' means second pass in case of time jumping back for one hour.

Now

<D/DTn>  = D/DT.today()                     # Current local date or naive datetime.
<DTn>    = DT.utcnow()                      # Naive datetime from current UTC time.
<DTa>    = DT.now(<tz>)                     # Aware datetime from current tz time.

Timezone

<tz>     = UTC                              # UTC timezone.
<tz>     = tzlocal()                        # Local timezone.
<tz>     = gettz('<Cont.>/<City>')          # Timezone from 'Continent/City_Name' str.

<DTa>    = <DT>.astimezone(<tz>)            # Datetime, converted to passed timezone.
<Ta/DTa> = <T/DT>.replace(tzinfo=<tz>)      # Unconverted object with new timezone.

Encode

<D/T/DT> = D/T/DT.fromisoformat('<iso>')    # Object from ISO string.
<DT>     = DT.strptime(<str>, '<format>')   # Datetime from str, according to format.
<D/DTn>  = D/DT.fromordinal(<int>)          # D/DTn from days since Christ.
<DTa>    = DT.fromtimestamp(<real>, <tz>)   # DTa from seconds since Epoch in tz time.

ISO strings come in following forms: 'YYYY-MM-DD', 'HH:MM:SS.ffffff[±<offset>]', or both separated by 'T'.
On Unix systems Epoch is '1970-01-01 00:00 UTC', '1970-01-01 01:00 CET', ...

Decode

<str>    = <D/T/DT>.isoformat()             # ISO string representation.
<str>    = <D/T/DT>.strftime('<format>')    # Custom string representation.
<int>    = <D/DT>.toordinal()               # Days since Christ, ignoring time and tz.
<float>  = <DT>.timestamp()                 # Seconds since Epoch in local time or tz.

Format

>>> dt = datetime.strptime('2015-05-14 23:39:00.00 +0200', '%Y-%m-%d %H:%M:%S.%f %z')
>>> dt.strftime("%A %dth %B '%y, %I:%M%p %Z")
"Thursday 14th May '15, 11:39PM UTC+02:00"

Rest of the codes:

'a' - Weekday, abbreviated name.
'b' - Month, abbreviated name.

Arguments

Inside Function Call

<function>(<positional_args>)                  # f(0, 0)
<function>(<keyword_args>)                     # f(x=0, y=0)
<function>(<positional_args>, <keyword_args>)  # f(0, y=0)

Inside Function Definition

def f(<nondefault_args>):                      # def f(x, y)
def f(<default_args>):                         # def f(x=0, y=0)
def f(<nondefault_args>, <default_args>):      # def f(x, y=0)

Splat Operator

Inside Function Call

Splat expands a collection into positional arguments, while splatty-splat expands a dictionary into keyword arguments.

args   = (1, 2)
kwargs = {'x': 3, 'y': 4, 'z': 5}
func(*args, **kwargs)

Is the same as:

func(1, 2, x=3, y=4, z=5)

Inside Function Definition

Splat combines zero or more positional arguments into a tuple, while splatty-splat combines zero or more keyword arguments into a dictionary.

def add(*a):
    return sum(a)

>>> add(1, 2, 3)
6

Legal argument combinations:

def f(*args):                  # f(1, 2, 3)
def f(x, *args):               # f(1, 2, 3)
def f(*args, z):               # f(1, 2, z=3)
def f(x, *args, z):            # f(1, 2, z=3)

def f(**kwargs):               # f(x=1, y=2, z=3)
def f(x, **kwargs):            # f(x=1, y=2, z=3) | f(1, y=2, z=3)

def f(*args, **kwargs):        # f(x=1, y=2, z=3) | f(1, y=2, z=3) | f(1, 2, z=3) | f(1, 2, 3)
def f(x, *args, **kwargs):     # f(x=1, y=2, z=3) | f(1, y=2, z=3) | f(1, 2, z=3) | f(1, 2, 3)
def f(*args, y, **kwargs):     # f(x=1, y=2, z=3) | f(1, y=2, z=3)
def f(x, *args, z, **kwargs):  # f(x=1, y=2, z=3) | f(1, y=2, z=3) | f(1, 2, z=3)

Other Uses

<list>  = [*<collection> [, ...]]
<set>   = {*<collection> [, ...]}
<tuple> = (*<collection>, [...])
<dict>  = {**<dict> [, ...]}

head, *body, tail = <collection>

Inline

Lambda

<function> = lambda: <return_value>
<function> = lambda <argument_1>, <argument_2>: <return_value>

Comprehension

<list> = [i+1 for i in range(10)]         # [1, 2, ..., 10]
<set>  = {i for i in range(10) if i > 5}  # {6, 7, 8, 9}
<iter> = (i+5 for i in range(10))         # (5, 6, ..., 14)
<dict> = {i: i*2 for i in range(10)}      # {0: 0, 1: 2, ..., 9: 18}

out = [i+j for i in range(10) for j in range(10)]

Is the same as:

out = []
for i in range(10):
    for j in range(10):
        out.append(i+j)

Map, Filter, Reduce

from functools import reduce
<iter> = map(lambda x: x + 1, range(10))            # (1, 2, ..., 10)
<iter> = filter(lambda x: x > 5, range(10))         # (6, 7, 8, 9)
<int>  = reduce(lambda out, x: out + x, range(10))  # 45

Any, All

<bool> = any(<collection>)                  # False if empty.
<bool> = all(el[1] for el in <collection>)  # True if empty.

If - Else

<expression_if_true> if <condition> else <expression_if_false>

>>> [a if a else 'zero' for a in (0, 1, 0, 3)]
['zero', 1, 'zero', 3]

Namedtuple, Enum, Class

from collections import namedtuple
Point     = namedtuple('Point', 'x y')
point     = Point(0, 0)

from enum import Enum
Direction = Enum('Direction', 'n e s w')
Cutlery   = Enum('Cutlery', {'fork': 1, 'knife': 2, 'spoon': 3})

# Warning: Objects will share the objects that are initialized in the dictionary!
Creature  = type('Creature', (), {'p': Point(0, 0), 'd': Direction.n})
creature  = Creature()

Closure

We have a closure in Python when:

A nested function references a value of its enclosing function and then
the enclosing function returns the nested function.

def get_multiplier(a):
    def out(b):
        return a * b
    return out

>>> multiply_by_3 = get_multiplier(3)
>>> multiply_by_3(10)
30

If multiple nested functions within enclosing function reference the same value, that value gets shared.
To dynamically access function's first free variable use '<function>.__closure__[0].cell_contents'.

Partial

from functools import partial
<function> = partial(<function> [, <arg_1>, <arg_2>, ...])

>>> import operator as op
>>> multiply_by_3 = partial(op.mul, 3)
>>> multiply_by_3(10)
30

Nonlocal

If variable is being assigned to anywhere in the scope, it is regarded as a local variable, unless it is declared as a 'global' or 'nonlocal'.

def get_counter():
    i = 0
    def out():
        nonlocal i
        i += 1
        return i
    return out

>>> counter = get_counter()
>>> counter(), counter(), counter()
(1, 2, 3)

Decorator

A decorator takes a function, adds some functionality and returns it.

@decorator_name
def function_that_gets_passed_to_decorator():
    ...

Debugger Example

Decorator that prints function's name every time it gets called.

from functools import wraps

def debug(func):
    @wraps(func)
    def out(*args, **kwargs):
        print(func.__name__)
        return func(*args, **kwargs)
    return out

@debug
def add(x, y):
    return x + y

Wraps is a helper decorator that copies metadata of function add() to function out().
Without it 'add.__name__' would return 'out'.

LRU Cache

Decorator that caches function's return values. All function's arguments must be hashable.

from functools import lru_cache

@lru_cache(maxsize=None)
def fib(n):
    return n if n < 2 else fib(n-2) + fib(n-1)

Recursion depth is limited to 1000 by default. To increase it use 'sys.setrecursionlimit(<depth>)'.

Parametrized Decorator

A decorator that accepts arguments and returns a normal decorator that accepts a function.

from functools import wraps

def debug(print_result=False):
    def decorator(func):
        @wraps(func)
        def out(*args, **kwargs):
            result = func(*args, **kwargs)
            print(func.__name__, result if print_result else '')
            return result
        return out
    return decorator

@debug(print_result=True)
def add(x, y):
    return x + y

Class

class <name>:
    def __init__(self, a):
        self.a = a
    def __repr__(self):
        class_name = self.__class__.__name__
        return f'{class_name}({self.a!r})'
    def __str__(self):
        return str(self.a)

    @classmethod
    def get_class_name(cls):
        return cls.__name__

Constructor Overloading

class <name>:
    def __init__(self, a=None):
        self.a = a

Inheritance

class Person:
    def __init__(self, name, age):
        self.name = name
        self.age  = age

class Employee(Person):
    def __init__(self, name, age, staff_num):
        super().__init__(name, age)
        self.staff_num = staff_num

>>> Employee.mro()
[<class 'Employee'>, <class 'Person'>, <class 'object'>]

Multiple Inheritance

class A: pass
class B: pass
class C(A, B): pass

MRO determines the order in which parent classes are traversed when searching for a method:

>>> C.mro()
[<class 'C'>, <class 'A'>, <class 'B'>, <class 'object'>]

Comparable

If eq() method is not overridden, it returns 'id(self) == id(other)', which is the same as 'self is other'.
That means all objects compare not equal by default.

class MyComparable:
    def __init__(self, a):
        self.a = a
    def __eq__(self, other):
        if isinstance(other, type(self)):
            return self.a == other.a
        return False

Hashable

Hashable object needs both hash() and eq() methods and its hash value should never change.
Hashable objects that compare equal must have the same hash value, meaning default hash() that returns 'id(self)' will not do.
That is why Python automatically makes classes unhashable if you only implement eq().

class MyHashable:
    def __init__(self, a):
        self.__a = copy.deepcopy(a)
    @property
    def a(self):
        return self.__a
    def __eq__(self, other):
        if isinstance(other, type(self)):
            return self.a == other.a
        return False
    def __hash__(self):
        return hash(self.a)

Sequence

Methods do not depend on each other, so they can be skipped if not needed.
Any object with defined getitem() is considered iterable, even if it lacks iter().

class MySequence:
    def __init__(self, a):
        self.a = a
    def __len__(self):
        return len(self.a)
    def __getitem__(self, i):
        return self.a[i]
    def __setitem__(self, i, value):
        self.a[i] = value
    def __iter__(self):
        for el in self.a:
            yield el

Callable

class Counter:
    def __init__(self):
        self.i = 0
    def __call__(self):
        self.i += 1
        return self.i

>>> counter = Counter()
>>> counter(), counter(), counter()
(1, 2, 3)

Withable

class MyOpen():
    def __init__(self, filename):
        self.filename = filename
    def __enter__(self):
        self.file = open(self.filename)
        return self.file
    def __exit__(self, *args):
        self.file.close()

>>> with open('test.txt', 'w') as file:
...     file.write('Hello World!')
>>> with MyOpen('test.txt') as file:
...     print(file.read())
Hello World!

Copy

from copy import copy, deepcopy
<object> = copy(<object>)
<object> = deepcopy(<object>)

Enum

from enum import Enum, auto

class <enum_name>(Enum):
    <member_name_1> = <value_1>
    <member_name_2> = <value_2_a>, <value_2_b>
    <member_name_3> = auto()

    @classmethod
    def get_member_names(cls):
        return [a.name for a in cls.__members__.values()]

<member> = <enum>.<member_name>
<member> = <enum>['<member_name>']
<member> = <enum>(<value>)
name     = <member>.name
value    = <member>.value

list_of_members = list(<enum>)
member_names    = [a.name for a in <enum>]
member_values   = [a.value for a in <enum>]
random_member   = random.choice(list(<enum>))

Inline

Cutlery = Enum('Cutlery', ['fork', 'knife', 'spoon'])
Cutlery = Enum('Cutlery', 'fork knife spoon')
Cutlery = Enum('Cutlery', {'fork': 1, 'knife': 2, 'spoon': 3})

Functions can not be values, so they must be wrapped:

from functools import partial
LogicOp = Enum('LogicOp', {'AND': partial(lambda l, r: l and r),
                           'OR' : partial(lambda l, r: l or r)})

Exceptions

while True:
    try:
        x = int(input('Please enter a number: '))
    except ValueError:
        print('Oops!  That was no valid number.  Try again...')
    else:
        print('Thank you.')
        break

Raising Exception

raise ValueError('A very specific message!')

Finally

>>> try:
...     raise KeyboardInterrupt
... finally:
...     print('Goodbye, world!')
Goodbye, world!
Traceback (most recent call last):
  File "<stdin>", line 2, in <module>
KeyboardInterrupt

Print

print(<el_1>, ..., sep=' ', end='\n', file=sys.stdout, flush=False)

Use 'file=sys.stderr' for errors.

Pretty Print

>>> from pprint import pprint
>>> pprint(dir())
['__annotations__',
 '__builtins__',
 '__doc__', ...]

Input

Reads a line from user input or pipe if present.
Trailing newline gets stripped.
Prompt string is printed to the standard output before reading input.

<str> = input(prompt=None)

Prints lines until EOF:

while True:
    try:
        print(input())
    except EOFError:
        break

Command Line Arguments

import sys
script_name = sys.argv[0]
arguments   = sys.argv[1:]

Argparse

from argparse import ArgumentParser, FileType
p = ArgumentParser(description=<str>)
p.add_argument('-<short_name>', '--<name>', action='store_true')  # Flag
p.add_argument('-<short_name>', '--<name>', type=<type>)          # Option
p.add_argument('<name>', type=<type>, nargs=1)                    # Argument
p.add_argument('<name>', type=<type>, nargs='+')                  # Arguments
args  = p.parse_args()
value = args.<name>

Use 'help=<str>' for argument description.
Use 'type=FileType(<mode>)' for files.

Open

Opens a file and returns a corresponding file object.

<file> = open('<path>', mode='r', encoding=None)

Modes

'r' - Read (default).
'w' - Write (truncate).
'x' - Write or fail if the file already exists.
'a' - Append.
'w+' - Read and write (truncate).
'r+' - Read and write from the start.
'a+' - Read and write from the end.
't' - Text mode (default).
'b' - Binary mode.

File

<file>.seek(0)                      # Moves to the start of the file.
<file>.seek(offset)                 # Moves 'offset' chars/bytes from the start.
<file>.seek(offset, <anchor>)       # Anchor: 0 start, 1 current pos., 2 end.

<str/bytes> = <file>.read(size=-1)  # Reads 'size' chars/bytes or until EOF.
<str/bytes> = <file>.readline()     # Returns a line.
<list>      = <file>.readlines()    # Returns a list of lines.
<str/bytes> = next(<file>)          # Returns a line using buffer. Do not mix.

<file>.write(<str/bytes>)           # Writes a string or bytes object.
<file>.writelines(<list>)           # Writes a list of strings or bytes objects.
<file>.flush()                      # Flushes write buffer.

Methods do not add or strip trailing newlines.

Read Text from File

def read_file(filename):
    with open(filename, encoding='utf-8') as file:
        return file.readlines()

Write Text to File

def write_to_file(filename, text):
    with open(filename, 'w', encoding='utf-8') as file:
        file.write(text)

Path

from os import path, listdir
<bool> = path.exists('<path>')
<bool> = path.isfile('<path>')
<bool> = path.isdir('<path>')
<list> = listdir('<path>')

>>> from glob import glob
>>> glob('../*.gif')
['1.gif', 'card.gif']

Pathlib

from pathlib import Path
cwd    = Path()
<Path> = Path('<path>' [, '<path>', <Path>, ...])
<Path> = <Path> / '<dir>' / '<file>'

<bool> = <Path>.exists()
<bool> = <Path>.is_file()
<bool> = <Path>.is_dir()
<iter> = <Path>.iterdir()

<iter> = <Path>.glob('<pattern>')

<str>  = str(<Path>)               # Returns path as a string.
<tup.> = <Path>.parts              # Returns all components as strings.
<Path> = <Path>.resolve()          # Returns absolute Path without symlinks.

<str>  = <Path>.name               # Final component.
<str>  = <Path>.stem               # Final component without extension.
<str>  = <Path>.suffix             # Final component's extension.
<Path> = <Path>.parent             # Path without final component.

Command Execution

import os
<str> = os.popen(<command>).read()

Subprocess

>>> import subprocess
>>> a = subprocess.run(['ls', '-a'], stdout=subprocess.PIPE)
>>> a.stdout
b'.\n..\nfile1.txt\nfile2.txt\n'
>>> a.returncode
0

CSV

import csv

Read Rows from CSV File

def read_csv_file(filename):
    with open(filename, encoding='utf-8') as file:
        return csv.reader(file, delimiter=';')

Write Rows to CSV File

def write_to_csv_file(filename, rows):
    with open(filename, 'w', encoding='utf-8') as file:
        writer = csv.writer(file, delimiter=';')
        writer.writerows(rows)

JSON

import json
<str>    = json.dumps(<object>, ensure_ascii=True, indent=None)
<object> = json.loads(<str>)

Read Object from JSON File

def read_json_file(filename):
    with open(filename, encoding='utf-8') as file:
        return json.load(file)

Write Object to JSON File

def write_to_json_file(filename, an_object):
    with open(filename, 'w', encoding='utf-8') as file:
        json.dump(an_object, file, ensure_ascii=False, indent=2)

Pickle

import pickle
<bytes>  = pickle.dumps(<object>)
<object> = pickle.loads(<bytes>)

Read Object from File

def read_pickle_file(filename):
    with open(filename, 'rb') as file:
        return pickle.load(file)

Write Object to File

def write_to_pickle_file(filename, an_object):
    with open(filename, 'wb') as file:
        pickle.dump(an_object, file)

SQLite

import sqlite3
db = sqlite3.connect('<path>')
...
db.close()

Read

cursor = db.execute('<query>')
if cursor:
    <tuple> = cursor.fetchone()  # First row.
    <list>  = cursor.fetchall()  # Remaining rows.

Write

db.execute('<query>')
db.commit()

Bytes

Bytes object is an immutable sequence of single bytes. Mutable version is called 'bytearray'.

<bytes> = b'<str>'
<int>   = <bytes>[<index>]
<bytes> = <bytes>[<slice>]
<ints>  = list(<bytes>)
<bytes> = b''.join(<coll_of_bytes>)

Encode

<bytes> = <str>.encode(encoding='utf-8')
<bytes> = <int>.to_bytes(<length>, byteorder='big|little', signed=False)
<bytes> = bytes.fromhex('<hex>')

Decode

<str>   = <bytes>.decode(encoding='utf-8')
<int>   = int.from_bytes(<bytes>, byteorder='big|little', signed=False)
<hex>   = <bytes>.hex()

Read Bytes from File

def read_bytes(filename):
    with open(filename, 'rb') as file:
        return file.read()

Write Bytes to File

def write_bytes(filename, bytes_obj):
    with open(filename, 'wb') as file:
        file.write(bytes_obj)

Struct

Module that performs conversions between Python values and a C struct, represented as a Python bytes object.
Machine’s native type sizes and byte order are used by default.

from struct import pack, unpack, iter_unpack, calcsize
<bytes>  = pack('<format>', <value_1> [, <value_2>, ...])
<tuple>  = unpack('<format>', <bytes>)
<tuples> = iter_unpack('<format>', <bytes>)

Example

>>> pack('>hhl', 1, 2, 3)
b'\x00\x01\x00\x02\x00\x00\x00\x03'
>>> unpack('>hhl', b'\x00\x01\x00\x02\x00\x00\x00\x03')
(1, 2, 3)
>>> calcsize('>hhl')
8

Format

For standard sizes start format string with:

'=' - native byte order
'<' - little-endian
'>' - big-endian

Use capital letter for unsigned type. Standard sizes are in brackets:

'x' - pad byte
'c' - char (1)
'h' - short (2)
'i' - int (4)
'l' - long (4)
'q' - long long (8)
'f' - float (4)
'd' - double (8)

Array

List that can hold only elements of predefined type. Available types are listed above.

from array import array
<array> = array('<typecode>' [, <collection>])

Memory View

Used for accessing the internal data of an object that supports the buffer protocol.

<memoryview> = memoryview(<bytes> / <bytearray> / <array>)
<memoryview>.release()

Deque

Thread-safe list with efficient appends and pops from either side. Pronounced "deck".

from collections import deque
<deque> = deque(<collection>, maxlen=None)

<deque>.appendleft(<el>)
<el> = <deque>.popleft()

<deque>.extendleft(<collection>)  # Collection gets reversed.
<deque>.rotate(n=1)               # Rotates elements to the right.

Threading

from threading import Thread, RLock

Thread

thread = Thread(target=<function>, args=(<first_arg>, ))
thread.start()
...
thread.join()

Lock

lock = RLock()
lock.acquire()
...
lock.release()

Introspection

Inspecting code at runtime.

Variables

<list> = dir()      # Names of variables in current scope.
<dict> = locals()   # Dict of local variables. Also vars().
<dict> = globals()  # Dict of global variables.

Attributes

<dict> = vars(<object>)
<bool> = hasattr(<object>, '<attr_name>')
value  = getattr(<object>, '<attr_name>')
setattr(<object>, '<attr_name>', value)

Parameters

from inspect import signature
<sig>        = signature(<function>)
no_of_params = len(<sig>.parameters)
param_names  = list(<sig>.parameters.keys())

Metaprograming

Code that generates code.

Type

Type is the root class. If only passed the object it returns its type (class). Otherwise it creates a new class.

<class> = type(<class_name>, <parents_tuple>, <attributes_dict>)

>>> Z = type('Z', (), {'a': 'abcde', 'b': 12345})
>>> z = Z()

Meta Class

Class that creates class.

def my_meta_class(name, parents, attrs):
    attrs['a'] = 'abcde'
    return type(name, parents, attrs)

Or:

class MyMetaClass(type):
    def __new__(cls, name, parents, attrs):
        attrs['a'] = 'abcde'
        return type.__new__(cls, name, parents, attrs)

New() is a class method that gets called before init(). If it returns an instance of its class, then that instance gets passed to init() as a 'self' argument.
It receives the same arguments as init(), except for the first one that specifies the desired class of returned instance ('MyMetaClass' in our case).
New() can also be called directly, usually from a new() method of a child class (def __new__(cls): return super().__new__(cls)), in which case init() is not called.

Metaclass Attribute

When class is created it checks if it has metaclass defined. If not, it recursively checks if any of his parents has it defined and eventually comes to type.

class MyClass(metaclass=MyMetaClass):
    b = 12345

>>> MyClass.a, MyClass.b
('abcde', 12345)

Operator

from operator import add, sub, mul, truediv, floordiv, mod, pow, neg, abs
from operator import eq, ne, lt, le, gt, ge
from operator import not_, and_, or_
from operator import itemgetter, attrgetter, methodcaller

import operator as op
product_of_elems = functools.reduce(op.mul, <collection>)
sorted_by_second = sorted(<collection>, key=op.itemgetter(1))
sorted_by_both   = sorted(<collection>, key=op.itemgetter(1, 0))
LogicOp          = enum.Enum('LogicOp', {'AND': op.and_, 'OR' : op.or_})
last_el          = op.methodcaller('pop')(<list>)

Eval

Basic

>>> from ast import literal_eval
>>> literal_eval('1 + 2')
3
>>> literal_eval('[1, 2, 3]')
[1, 2, 3]
>>> literal_eval('abs(1)')
ValueError: malformed node or string

Using Abstract Syntax Trees

import ast
from ast import Num, BinOp, UnaryOp
import operator as op

LEGAL_OPERATORS = {ast.Add:    op.add,      # <el> + <el>
                   ast.Sub:    op.sub,      # <el> - <el>
                   ast.Mult:   op.mul,      # <el> * <el>
                   ast.Div:    op.truediv,  # <el> / <el>
                   ast.Pow:    op.pow,      # <el> ** <el>
                   ast.BitXor: op.xor,      # <el> ^ <el>
                   ast.USub:   op.neg}      # - <el>

def evaluate(expression):
    root = ast.parse(expression, mode='eval')
    return eval_node(root.body)

def eval_node(node):
    node_type = type(node)
    if node_type == Num:
        return node.n
    if node_type not in [BinOp, UnaryOp]:
        raise TypeError(node)
    operator_type = type(node.op)
    if operator_type not in LEGAL_OPERATORS:
        raise TypeError(f'Illegal operator {node.op}')
    operator = LEGAL_OPERATORS[operator_type]
    if node_type == BinOp:
        left, right = eval_node(node.left), eval_node(node.right)
        return operator(left, right)
    elif node_type == UnaryOp:
        operand = eval_node(node.operand)
        return operator(operand)

>>> evaluate('2 ^ 6')
4
>>> evaluate('2 ** 6')
64
>>> evaluate('1 + 2 * 3 ** (4 ^ 5) / (6 + -7)')
-5.0

Coroutine

Similar to generator, but generator pulls data through the pipe with iteration, while coroutine pushes data into the pipeline with send().
Coroutines provide more powerful data routing possibilities than iterators.
If you build a collection of simple data processing components, you can glue them together into complex arrangements of pipes, branches, merging, etc.

Helper Decorator

All coroutines must be "primed" by first calling next().
Remembering to call next() is easy to forget.
Solved by wrapping coroutines with a decorator:

def coroutine(func):
    def out(*args, **kwargs):
        cr = func(*args, **kwargs)
        next(cr)
        return cr
    return out

Pipeline Example

def reader(target):
    for i in range(10):
        target.send(i)
    target.close()

@coroutine
def adder(target):
    while True:
        value = (yield)
        target.send(value + 100)

@coroutine
def printer():
    while True:
        value = (yield)
        print(value)

reader(adder(printer()))  # 100, 101, ..., 109

Libraries

Progress Bar

# $ pip3 install tqdm
from tqdm import tqdm
from time import sleep
for i in tqdm([1, 2, 3]):
    sleep(0.2)
for i in tqdm(range(100)):
    sleep(0.02)

Plot

# $ pip3 install matplotlib
from matplotlib import pyplot
pyplot.plot(<data_1> [, <data_2>, ...])
pyplot.savefig(<filename>, transparent=True)
pyplot.show()

Table

Prints CSV file as ASCII table:

# $ pip3 install tabulate
from tabulate import tabulate
import csv
with open(<filename>, encoding='utf-8') as file:
    lines   = csv.reader(file, delimiter=';')
    headers = [header.title() for header in next(lines)]
    table   = tabulate(lines, headers)
    print(table)

Curses

from curses import wrapper

def main():
    wrapper(draw)

def draw(screen):
    screen.clear()
    screen.addstr(0, 0, 'Press ESC to quit.')
    while screen.getch() != 27:
        pass

def get_border(screen):
    from collections import namedtuple
    P = namedtuple('P', 'x y')
    height, width = screen.getmaxyx()
    return P(width-1, height-1)

Logging

# $ pip3 install loguru
from loguru import logger

logger.add('debug_{time}.log', colorize=True)  # Connects a log file.
logger.add('error_{time}.log', level='ERROR')  # Another file for errors or higher.
logger.<level>('A logging message')

Levels: 'debug', 'info', 'success', 'warning', 'error', 'critical'.

Rotation

Parameter that sets a condition when a new log file is created.

rotation=<int>|<datetime.timedelta>|<datetime.time>|<str>

'<int>' - Max file size in bytes.
'<timedelta>' - Max age of a file.
'<time>' - Time of day.
'<str>' - Any of above as a string: '100 MB', '1 month', 'monday at 12:00', ...

Retention

Sets a condition which old log files are deleted.

retention=<int>|<datetime.timedelta>|<str>

'<int>' - Max number of files.
'<timedelta>' - Max age of a file.
'<str>' - Max age as a string: '1 week, 3 days', '2 months', ...

Compression

Sets how inactive log files are compressed.

compression='gz'|'bz2'|'tar'|'tar.gz'|'tar.bz2'|'zip'

Scraping

# $ pip3 install requests beautifulsoup4
>>> import requests
>>> from bs4 import BeautifulSoup
>>> url   = 'https://en.wikipedia.org/wiki/Python_(programming_language)'
>>> page  = requests.get(url)
>>> doc   = BeautifulSoup(page.text, 'html.parser')
>>> table = doc.find('table', class_='infobox vevent')
>>> rows  = table.find_all('tr')
>>> link  = rows[11].find('a')['href']
>>> ver   = rows[6].find('div').text.split()[0]
>>> link, ver
('https://www.python.org/', '3.7.2')

Web

# $ pip3 install bottle
from bottle import run, route, post, template, request, response
import json

Run

run(host='localhost', port=8080)
run(host='0.0.0.0', port=80, server='cherrypy')

Static Request

@route('/img/<image>')
def send_image(image):
    return static_file(image, 'images/', mimetype='image/png')

Dynamic Request

@route('/<sport>')
def send_page(sport):
    return template('<h1>{{title}}</h1>', title=sport)

REST Request

@post('/odds/<sport>')
def odds_handler(sport):
    team = request.forms.get('team')
    home_odds, away_odds = 2.44, 3.29
    response.headers['Content-Type'] = 'application/json'
    response.headers['Cache-Control'] = 'no-cache'
    return json.dumps([team, home_odds, away_odds])

Test:

# $ pip3 install requests
>>> import requests
>>> url  = 'http://localhost:8080/odds/football'
>>> data = {'team': 'arsenal f.c.'}
>>> response = requests.post(url, data=data)
>>> response.json()
['arsenal f.c.', 2.44, 3.29]

Profile

Basic

from time import time
start_time = time()  # Seconds since Epoch.
...
duration = time() - start_time

High Performance

from time import perf_counter as pc
start_time = pc()    # Seconds since restart.
...
duration = pc() - start_time

Timing a Snippet

from timeit import timeit
timeit('"-".join(str(a) for a in range(100))', 
       number=10000, globals=globals(), setup='pass')

Line Profiler

# $ pip3 install line_profiler
@profile
def main():
    a = [*range(10000)]
    b = {*range(10000)}
main()

Usage:

$ kernprof -lv test.py
Line #      Hits         Time  Per Hit   % Time  Line Contents
==============================================================
     1                                           @profile
     2                                           def main():
     3         1       1128.0   1128.0     27.4      a = [*range(10000)]
     4         1       2994.0   2994.0     72.6      b = {*range(10000)}

Call Graph

Generates a PNG image of a call graph with highlighted bottlenecks:

# $ pip3 install pycallgraph
from pycallgraph import output, PyCallGraph
from datetime import datetime
time_str = datetime.now().strftime('%Y%m%d%H%M%S')
filename = f'profile-{time_str}.png'
drawer = output.GraphvizOutput(output_file=filename)
with PyCallGraph(output=drawer):
    <code_to_be_profiled>

NumPy

Array manipulation mini language. Can run up to one hundred times faster than equivalent Python code.

# $ pip3 install numpy
import numpy as np

<array> = np.array(<list>)
<array> = np.arange(from_inclusive, to_exclusive, ±step_size)
<array> = np.ones(<shape>)
<array> = np.random.randint(from_inclusive, to_exclusive, <shape>)

<array>.shape = <shape>
<view>  = <array>.reshape(<shape>)
<view>  = np.broadcast_to(<array>, <shape>)

<array> = <array>.sum(axis)
indexes = <array>.argmin(axis)

Shape is a tuple of dimension sizes.
Axis is an index of dimension that gets collapsed. Leftmost dimension has index 0.

Indexing

<el>       = <2d_array>[0, 0]        # First element.
<1d_view>  = <2d_array>[0]           # First row.
<1d_view>  = <2d_array>[:, 0]        # First column. Also [..., 0].
<3d_view>  = <2d_array>[None, :, :]  # Expanded by dimension of size 1.

<1d_array> = <2d_array>[<1d_row_indexes>, <1d_column_indexes>]
<2d_array> = <2d_array>[<2d_row_indexes>, <2d_column_indexes>]

<2d_bools> = <2d_array> > 0
<1d_array> = <2d_array>[<2d_bools>]

If row and column indexes differ in shape, they are combined with broadcasting.

Broadcasting

Broadcasting is a set of rules by which NumPy functions operate on arrays of different sizes and/or dimensions.

left  = [[0.1], [0.6], [0.8]]  # Shape: (3, 1)
right = [ 0.1 ,  0.6 ,  0.8 ]  # Shape: (3)

1. If array shapes differ in length, left-pad the smaller shape with ones:

left  = [[0.1], [0.6], [0.8]]  # Shape: (3, 1)
right = [[0.1 ,  0.6 ,  0.8]]  # Shape: (1, 3) <- !

2. If any dimensions differ in size, expand the ones that have size 1 by duplicating their elements:

left  = [[0.1, 0.1, 0.1], [0.6, 0.6, 0.6], [0.8, 0.8, 0.8]]  # Shape: (3, 3) <- !
right = [[0.1, 0.6, 0.8], [0.1, 0.6, 0.8], [0.1, 0.6, 0.8]]  # Shape: (3, 3) <- !

3. If neither non-matching dimension has size 1, rise an error.

Example

For each point returns index of its nearest point (`[0.1, 0.6, 0.8] => [1, 2, 1]`):

>>> points = np.array([0.1, 0.6, 0.8])
[ 0.1,  0.6,  0.8]
>>> wrapped_points = points.reshape(3, 1)
[[ 0.1],
 [ 0.6],
 [ 0.8]]
>>> distances = wrapped_points - points
[[ 0. , -0.5, -0.7],
 [ 0.5,  0. , -0.2],
 [ 0.7,  0.2,  0. ]]
>>> distances = np.abs(distances)
[[ 0. ,  0.5,  0.7],
 [ 0.5,  0. ,  0.2],
 [ 0.7,  0.2,  0. ]]
>>> i = np.arange(3)
[0, 1, 2]
>>> distances[i, i] = np.inf
[[ inf,  0.5,  0.7],
 [ 0.5,  inf,  0.2],
 [ 0.7,  0.2,  inf]]
>>> distances.argmin(1)
[1, 2, 1]

Image

# $ pip3 install pillow
from PIL import Image

Creates PNG image of a rainbow gradient:

width  = 100
height = 100
size   = width * height
pixels = [255 * i/size for i in range(size)]

img = Image.new('HSV', (width, height))
img.putdata([(int(a), 255, 255) for a in pixels])
img.convert(mode='RGB').save('test.png')

Adds noise to a PNG image:

from random import randint
add_noise = lambda value: max(0, min(255, value + randint(-20, 20)))
img = Image.open('test.png').convert(mode='HSV')
img.putdata([(add_noise(h), s, v) for h, s, v in img.getdata()])
img.convert(mode='RGB').save('test.png')

Modes

'1' - 1-bit pixels, black and white, stored with one pixel per byte.
'L' - 8-bit pixels, greyscale.
'RGB' - 3x8-bit pixels, true color.
'RGBA' - 4x8-bit pixels, true color with transparency mask.
'HSV' - 3x8-bit pixels, Hue, Saturation, Value color space.

Audio

import wave
from struct import pack, iter_unpack

Read Frames from WAV File

def read_wav_file(filename):
    with wave.open(filename, 'rb') as wf:
        frames = wf.readframes(wf.getnframes())
        return [a[0] for a in iter_unpack('<h', frames)]

Write Frames to WAV File

def write_to_wav_file(filename, frames_int, mono=True):
    frames_short = (pack('<h', a) for a in frames_int)
    with wave.open(filename, 'wb') as wf:
        wf.setnchannels(1 if mono else 2)
        wf.setsampwidth(2)
        wf.setframerate(44100)
        wf.writeframes(b''.join(frames_short))

Examples

Saves a sine wave to a mono WAV file:

from math import pi, sin
frames_f = (sin(i * 2 * pi * 440 / 44100) for i in range(100000))
frames_i = (int(a * 30000) for a in frames_f)
write_to_wav_file('test.wav', frames_i)

Adds noise to a mono WAV file:

from random import randint
add_noise = lambda value: max(-32768, min(32767, value + randint(-500, 500)))
frames_i  = (add_noise(a) for a in read_wav_file('test.wav'))
write_to_wav_file('test.wav', frames_i)

Plays Popcorn:

# $ pip3 install simpleaudio
import simpleaudio, math, struct
from itertools import chain, repeat
F  = 44100
P1 = '71♪,69,,71♪,66,,62♪,66,,59♪,,,'
P2 = '71♪,73,,74♪,73,,74,,71,,73♪,71,,73,,69,,71♪,69,,71,,67,,71♪,,,'
get_pause = lambda seconds: repeat(0, int(seconds * F))
sin_f     = lambda i, hz: math.sin(i * 2 * math.pi * hz / F)
get_wave  = lambda hz, seconds: (sin_f(i, hz) for i in range(int(seconds * F)))
get_hz    = lambda key: 8.176 * 2 ** (int(key) / 12)
parse_n   = lambda note: (get_hz(note[:2]), 0.25 if '♪' in note else 0.125)
get_note  = lambda note: get_wave(*parse_n(note)) if note else get_pause(0.125)
frames_i  = chain.from_iterable(get_note(n) for n in f'{P1}{P1}{P2}'.split(','))
frames_b  = b''.join(struct.pack('<h', int(a * 30000)) for a in frames_i)
simpleaudio.play_buffer(frames_b, 1, 2, F)

Basic Script Template

#!/usr/bin/env python3
#
# Usage: .py
# 

from collections import namedtuple
from enum import Enum
import re
import sys


def main():
    pass


###
##  UTIL
#

def read_file(filename):
    with open(filename, encoding='utf-8') as file:
        return file.readlines()


if __name__ == '__main__':
    main()

hkyyx/python-cheatsheet

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Languages

Packages