What does ** (double star/asterisk) and * (star/asterisk) do for parameters?

The *args and **kwargs is a common idiom to allow arbitrary number of arguments to functions as described in the section more on defining functions in the Python documentation.

The *args will give you all function parameters as a tuple:

def foo(*args):
    for a in args:
        print(a)        

foo(1)
# 1

foo(1,2,3)
# 1
# 2
# 3

The **kwargs will give you all keyword arguments except for those corresponding to a formal parameter as a dictionary.

def bar(**kwargs):
    for a in kwargs:
        print(a, kwargs[a])  

bar(name='one', age=27)
# name one
# age 27

Both idioms can be mixed with normal arguments to allow a set of fixed and some variable arguments:

def foo(kind, *args, **kwargs):
   pass

It is also possible to use this the other way around:

def foo(a, b, c):
    print(a, b, c)

obj = {'b':10, 'c':'lee'}

foo(100,**obj)
# 100 10 lee

Another usage of the *l idiom is to unpack argument lists when calling a function.

def foo(bar, lee):
    print(bar, lee)

l = [1,2]

foo(*l)
# 1 2

In Python 3 it is possible to use *l on the left side of an assignment (Extended Iterable Unpacking), though it gives a list instead of a tuple in this context:

first, *rest = [1,2,3,4]
first, *l, last = [1,2,3,4]

Also Python 3 adds new semantic (refer PEP 3102):

def func(arg1, arg2, arg3, *, kwarg1, kwarg2):
    pass

For example the following works in python 3 but not python 2:

>>> x = [1, 2]
>>> [*x]
[1, 2]
>>> [*x, 3, 4]
[1, 2, 3, 4]

>>> x = {1:1, 2:2}
>>> x
{1: 1, 2: 2}
>>> {**x, 3:3, 4:4}
{1: 1, 2: 2, 3: 3, 4: 4}

Such function accepts only 3 positional arguments, and everything after * can only be passed as keyword arguments.

Note:

• A Python dict, semantically used for keyword argument passing, are arbitrarily ordered. However, in Python 3.6, keyword arguments are guaranteed to remember insertion order.
• "The order of elements in **kwargs now corresponds to the order in which keyword arguments were passed to the function." - What’s New In Python 3.6
• In fact, all dicts in CPython 3.6 will remember insertion order as an implementation detail, this becomes standard in Python 3.7.

It's also worth noting that you can use * and ** when calling functions as well. This is a shortcut that allows you to pass multiple arguments to a function directly using either a list/tuple or a dictionary. For example, if you have the following function:

def foo(x,y,z):
    print("x=" + str(x))
    print("y=" + str(y))
    print("z=" + str(z))

You can do things like:

>>> mylist = [1,2,3]
>>> foo(*mylist)
x=1
y=2
z=3

>>> mydict = {'x':1,'y':2,'z':3}
>>> foo(**mydict)
x=1
y=2
z=3

>>> mytuple = (1, 2, 3)
>>> foo(*mytuple)
x=1
y=2
z=3

Note: The keys in mydict have to be named exactly like the parameters of function foo. Otherwise it will throw a TypeError:

>>> mydict = {'x':1,'y':2,'z':3,'badnews':9}
>>> foo(**mydict)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: foo() got an unexpected keyword argument 'badnews'

The single * means that there can be any number of extra positional arguments. foo() can be invoked like foo(1,2,3,4,5). In the body of foo() param2 is a sequence containing 2-5.

The double ** means there can be any number of extra named parameters. bar() can be invoked like bar(1, a=2, b=3). In the body of bar() param2 is a dictionary containing {'a':2, 'b':3 }

With the following code:

def foo(param1, *param2):
    print(param1)
    print(param2)

def bar(param1, **param2):
    print(param1)
    print(param2)

foo(1,2,3,4,5)
bar(1,a=2,b=3)

the output is

1
(2, 3, 4, 5)
1
{'a': 2, 'b': 3}

For those of you who learn by examples!

1. The purpose of * is to give you the ability to define a function that can take an arbitrary number of arguments provided as a list (e.g. f(*myList) ).
2. The purpose of ** is to give you the ability to feed a function's arguments by providing a dictionary (e.g. f(**{'x' : 1, 'y' : 2}) ).

Let us show this by defining a function that takes two normal variables x, y, and can accept more arguments as myArgs, and can accept even more arguments as myKW. Later, we will show how to feed y using myArgDict.

def f(x, y, *myArgs, **myKW):
    print("# x      = {}".format(x))
    print("# y      = {}".format(y))
    print("# myArgs = {}".format(myArgs))
    print("# myKW   = {}".format(myKW))
    print("# ----------------------------------------------------------------------")

# Define a list for demonstration purposes
myList    = ["Left", "Right", "Up", "Down"]
# Define a dictionary for demonstration purposes
myDict    = {"Wubba": "lubba", "Dub": "dub"}
# Define a dictionary to feed y
myArgDict = {'y': "Why?", 'y0': "Why not?", "q": "Here is a cue!"}

# The 1st elem of myList feeds y
f("myEx", *myList, **myDict)
# x      = myEx
# y      = Left
# myArgs = ('Right', 'Up', 'Down')
# myKW   = {'Wubba': 'lubba', 'Dub': 'dub'}
# ----------------------------------------------------------------------

# y is matched and fed first
# The rest of myArgDict becomes additional arguments feeding myKW
f("myEx", **myArgDict)
# x      = myEx
# y      = Why?
# myArgs = ()
# myKW   = {'y0': 'Why not?', 'q': 'Here is a cue!'}
# ----------------------------------------------------------------------

# The rest of myArgDict becomes additional arguments feeding myArgs
f("myEx", *myArgDict)
# x      = myEx
# y      = y
# myArgs = ('y0', 'q')
# myKW   = {}
# ----------------------------------------------------------------------

# Feed extra arguments manually and append even more from my list
f("myEx", 4, 42, 420, *myList, *myDict, **myDict)
# x      = myEx
# y      = 4
# myArgs = (42, 420, 'Left', 'Right', 'Up', 'Down', 'Wubba', 'Dub')
# myKW   = {'Wubba': 'lubba', 'Dub': 'dub'}
# ----------------------------------------------------------------------

# Without the stars, the entire provided list and dict become x, and y:
f(myList, myDict)
# x      = ['Left', 'Right', 'Up', 'Down']
# y      = {'Wubba': 'lubba', 'Dub': 'dub'}
# myArgs = ()
# myKW   = {}
# ----------------------------------------------------------------------

Caveats

1. ** is exclusively reserved for dictionaries.
2. Non-optional argument assignment happens first.
3. You cannot use a non-optional argument twice.
4. If applicable, ** must come after *, always.

TL;DR

Below are 6 different use cases for * and ** in python programming:

1. To accept any number of positional arguments using *args: def foo(*args): pass, here foo accepts any number of positional arguments, i. e., the following calls are valid foo(1), foo(1, 'bar')
2. To accept any number of keyword arguments using **kwargs: def foo(**kwargs): pass, here 'foo' accepts any number of keyword arguments, i. e., the following calls are valid foo(name='Tom'), foo(name='Tom', age=33)
3. To accept any number of positional and keyword arguments using *args, **kwargs: def foo(*args, **kwargs): pass, here foo accepts any number of positional and keyword arguments, i. e., the following calls are valid foo(1,name='Tom'), foo(1, 'bar', name='Tom', age=33)
4. To enforce keyword only arguments using *: def foo(pos1, pos2, *, kwarg1): pass, here * means that foo only accept keyword arguments after pos2, hence foo(1, 2, 3) raises TypeError but foo(1, 2, kwarg1=3) is ok.
5. To express no further interest in more positional arguments using *_ (Note: this is a convention only): def foo(bar, baz, *_): pass means (by convention) foo only uses bar and baz arguments in its working and will ignore others.
6. To express no further interest in more keyword arguments using \**_ (Note: this is a convention only): def foo(bar, baz, **_): pass means (by convention) foo only uses bar and baz arguments in its working and will ignore others.

BONUS: From python 3.8 onward, one can use / in function definition to enforce positional only parameters. In the following example, parameters a and b are positional-only, while c or d can be positional or keyword, and e or f are required to be keywords:

def f(a, b, /, c, d, *, e, f):
    pass

From the Python documentation:

If there are more positional arguments than there are formal parameter slots, a TypeError exception is raised, unless a formal parameter using the syntax "*identifier" is present; in this case, that formal parameter receives a tuple containing the excess positional arguments (or an empty tuple if there were no excess positional arguments).

If any keyword argument does not correspond to a formal parameter name, a TypeError exception is raised, unless a formal parameter using the syntax "**identifier" is present; in this case, that formal parameter receives a dictionary containing the excess keyword arguments (using the keywords as keys and the argument values as corresponding values), or a (new) empty dictionary if there were no excess keyword arguments.

* means receive variable arguments as tuple

** means receive variable arguments as dictionary

Used like the following:

1) single *

def foo(*args):
    for arg in args:
        print(arg)

foo("two", 3)

Output:

two
3

2) Now **

def bar(**kwargs):
    for key in kwargs:
        print(key, kwargs[key])

bar(dic1="two", dic2=3)

Output:

dic1 two
dic2 3

TL;DR

It packs arguments passed to the function into list and dict respectively inside the function body. When you define a function signature like this:

def func(*args, **kwds):
    # do stuff

it can be called with any number of arguments and keyword arguments. The non-keyword arguments get packed into a list called args inside the function body and the keyword arguments get packed into a dict called kwds inside the function body.

func("this", "is a list of", "non-keyowrd", "arguments", keyword="ligma", options=[1,2,3])

now inside the function body, when the function is called, there are two local variables, args which is a list having value ["this", "is a list of", "non-keyword", "arguments"] and kwds which is a dict having value {"keyword" : "ligma", "options" : [1,2,3]}

This also works in reverse, i.e. from the caller side. for example if you have a function defined as:

def f(a, b, c, d=1, e=10):
    # do stuff

you can call it with by unpacking iterables or mappings you have in the calling scope:

iterable = [1, 20, 500]
mapping = {"d" : 100, "e": 3}
f(*iterable, **mapping)
# That call is equivalent to
f(1, 20, 500, d=100, e=3)

Building on nickd's answer...

def foo(param1, *param2):
    print(param1)
    print(param2)


def bar(param1, **param2):
    print(param1)
    print(param2)


def three_params(param1, *param2, **param3):
    print(param1)
    print(param2)
    print(param3)


foo(1, 2, 3, 4, 5)
print("\n")
bar(1, a=2, b=3)
print("\n")
three_params(1, 2, 3, 4, s=5)

Output:

1
(2, 3, 4, 5)

1
{'a': 2, 'b': 3}

1
(2, 3, 4)
{'s': 5}

Basically, any number of positional arguments can use *args and any named arguments (or kwargs aka keyword arguments) can use **kwargs.

*args and **kwargs: allow you to pass a variable number of arguments to a function.

*args: is used to send a non-keyworded variable length argument list to the function:

def args(normal_arg, *argv):
    print("normal argument:", normal_arg)

    for arg in argv:
        print("Argument in list of arguments from *argv:", arg)

args('animals', 'fish', 'duck', 'bird')

Will produce:

normal argument: animals
Argument in list of arguments from *argv: fish
Argument in list of arguments from *argv: duck
Argument in list of arguments from *argv: bird

**kwargs*

**kwargs allows you to pass keyworded variable length of arguments to a function. You should use **kwargs if you want to handle named arguments in a function.

def who(**kwargs):
    if kwargs is not None:
        for key, value in kwargs.items():
            print("Your %s is %s." % (key, value))

who(name="Nikola", last_name="Tesla", birthday="7.10.1856", birthplace="Croatia")  

Will produce:

Your name is Nikola.
Your last_name is Tesla.
Your birthday is 7.10.1856.
Your birthplace is Croatia.

Given a function that has 3 items as argument

sum = lambda x, y, z: x + y + z
sum(1,2,3) # sum 3 items

sum([1,2,3]) # error, needs 3 items, not 1 list

x = [1,2,3][0]
y = [1,2,3][1]
z = [1,2,3][2]
sum(x,y,z) # ok

sum(*[1,2,3]) # ok, 1 list becomes 3 items

Imagine this toy with a bag of a triangle, a circle and a rectangle item. That bag does not directly fit. You need to unpack the bag to take those 3 items and now they fit. The Python * operator does this unpack process.

A good example of using both in a function is:

>>> def foo(*arg,**kwargs):
...     print arg
...     print kwargs
>>>
>>> a = (1, 2, 3)
>>> b = {'aa': 11, 'bb': 22}
>>>
>>>
>>> foo(*a,**b)
(1, 2, 3)
{'aa': 11, 'bb': 22}
>>>
>>>
>>> foo(a,**b) 
((1, 2, 3),)
{'aa': 11, 'bb': 22}
>>>
>>>
>>> foo(a,b) 
((1, 2, 3), {'aa': 11, 'bb': 22})
{}
>>>
>>>
>>> foo(a,*b)
((1, 2, 3), 'aa', 'bb')
{}

Context

• python 3.x
• unpacking with **
• use with string formatting

Use with string formatting

In addition to the answers in this thread, here is another detail that was not mentioned elsewhere. This expands on the answer by Brad Solomon

Unpacking with ** is also useful when using python str.format.

This is somewhat similar to what you can do with python f-strings f-string but with the added overhead of declaring a dict to hold the variables (f-string does not require a dict).

Quick Example

  ## init vars
  ddvars = dict()
  ddcalc = dict()
  pass
  ddvars['fname']     = 'Huomer'
  ddvars['lname']     = 'Huimpson'
  ddvars['motto']     = 'I love donuts!'
  ddvars['age']       = 33
  pass
  ddcalc['ydiff']     = 5
  ddcalc['ycalc']     = ddvars['age'] + ddcalc['ydiff']
  pass
  vdemo = []

  ## ********************
  ## single unpack supported in py 2.7
  vdemo.append('''
  Hello {fname} {lname}!

  Today you are {age} years old!

  We love your motto "{motto}" and we agree with you!
  '''.format(**ddvars)) 
  pass

  ## ********************
  ## multiple unpack supported in py 3.x
  vdemo.append('''
  Hello {fname} {lname}!

  In {ydiff} years you will be {ycalc} years old!
  '''.format(**ddvars,**ddcalc)) 
  pass

  ## ********************
  print(vdemo[-1])

*args ( or *any ) means every parameters

def any_param(*param):
    pass

any_param(1)
any_param(1,1)
any_param(1,1,1)
any_param(1,...)

NOTICE : you can don't pass parameters to *args

def any_param(*param):
    pass

any_param() # will work correct

The *args is in type tuple

def any_param(*param):
    return type(param)

any_param(1) #tuple
any_param() # tuple

for access to elements don't use of *

def any(*param):
    param[0] # correct

def any(*param):
    *param[0] # incorrect

The **kwd

**kwd or **any This is a dict type

def func(**any):
    return type(any) # dict

def func(**any):
    return any

func(width="10",height="20") # {width="10",height="20")

• def foo(param1, *param2): is a method can accept arbitrary number of values for *param2,
• def bar(param1, **param2): is a method can accept arbitrary number of values with keys for *param2
• param1 is a simple parameter.

For example, the syntax for implementing varargs in Java as follows:

accessModifier methodName(datatype… arg) {
    // method body
}

"Infinite" Args with *args and **kwargs

*args and **kwargs are just some way to input unlimited characters to functions, like:

def print_all(*args, **kwargs):
    print(args) # print any number of arguments like: "print_all("foo", "bar")"
    print(kwargs.get("to_print")) # print the value of the keyworded argument "to_print"


# example:
print_all("Hello", "World", to_print="!")
# will print:
"""
('Hello', 'World')
!
"""