# Python Dictionary Comprehension

## The Question :

411 people think this question is useful

Is it possible to create a dictionary comprehension in Python (for the keys)?

Without list comprehensions, you can use something like this:

l = []
for n in range(1, 11):
l.append(n)



We can shorten this to a list comprehension: l = [n for n in range(1, 11)].

However, say I want to set a dictionary’s keys to the same value. I can do:

d = {}
for n in range(1, 11):
d[n] = True # same value for each



I’ve tried this:

d = {}
d[i for i in range(1, 11)] = True



However, I get a SyntaxError on the for.

In addition (I don’t need this part, but just wondering), can you set a dictionary’s keys to a bunch of different values, like this:

d = {}
for n in range(1, 11):
d[n] = n



Is this possible with a dictionary comprehension?

d = {}
d[i for i in range(1, 11)] = [x for x in range(1, 11)]



This also raises a SyntaxError on the for.

• For future readers’ info: NumPy arrays do let you set multiple elements to a single value or list of values, the way you’re trying to do. Though if you don’t already have a reason to use NumPy, it’s probably not worth it just for this feature.

544 people think this answer is useful

There are dictionary comprehensions in Python 2.7+, but they don’t work quite the way you’re trying. Like a list comprehension, they create a new dictionary; you can’t use them to add keys to an existing dictionary. Also, you have to specify the keys and values, although of course you can specify a dummy value if you like.

>>> d = {n: n**2 for n in range(5)}
>>> print d
{0: 0, 1: 1, 2: 4, 3: 9, 4: 16}



If you want to set them all to True:

>>> d = {n: True for n in range(5)}
>>> print d
{0: True, 1: True, 2: True, 3: True, 4: True}



What you seem to be asking for is a way to set multiple keys at once on an existing dictionary. There’s no direct shortcut for that. You can either loop like you already showed, or you could use a dictionary comprehension to create a new dict with the new values, and then do oldDict.update(newDict) to merge the new values into the old dict.

152 people think this answer is useful

You can use the dict.fromkeys class method …

>>> dict.fromkeys(range(5), True)
{0: True, 1: True, 2: True, 3: True, 4: True}



This is the fastest way to create a dictionary where all the keys map to the same value.

d = dict.fromkeys(range(5), [])
# {0: [], 1: [], 2: [], 3: [], 4: []}
d.append(2)
# {0: , 1: , 2: , 3: , 4: } !!!



If you don’t actually need to initialize all the keys, a defaultdict might be useful as well:

from collections import defaultdict
d = defaultdict(True)



To answer the second part, a dict-comprehension is just what you need:

{k: k for k in range(10)}



You probably shouldn’t do this but you could also create a subclass of dict which works somewhat like a defaultdict if you override __missing__:

>>> class KeyDict(dict):
...    def __missing__(self, key):
...       #self[key] = key  # Maybe add this also?
...       return key
...
>>> d = KeyDict()
>>> d
1
>>> d
2
>>> d
3
>>> print(d)
{}



29 people think this answer is useful
>>> {i:i for i in range(1, 11)}
{1: 1, 2: 2, 3: 3, 4: 4, 5: 5, 6: 6, 7: 7, 8: 8, 9: 9, 10: 10}



25 people think this answer is useful

I really like the @mgilson comment, since if you have a two iterables, one that corresponds to the keys and the other the values, you can also do the following.

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



giving

d = {‘a’: 1, ‘b’: 2, ‘c’: 3}

12 people think this answer is useful

Use dict() on a list of tuples, this solution will allow you to have arbitrary values in each list, so long as they are the same length

i_s = range(1, 11)
x_s = range(1, 11)
# x_s = range(11, 1, -1) # Also works
d = dict([(i_s[index], x_s[index], ) for index in range(len(i_s))])



12 people think this answer is useful

Consider this example of counting the occurrence of words in a list using dictionary comprehension

my_list = ['hello', 'hi', 'hello', 'today', 'morning', 'again', 'hello']
my_dict = {k:my_list.count(k) for k in my_list}
print(my_dict)



And the result is

{'again': 1, 'hi': 1, 'hello': 3, 'today': 1, 'morning': 1}



9 people think this answer is useful

The main purpose of a list comprehension is to create a new list based on another one without changing or destroying the original list.

l = []
for n in range(1, 11):
l.append(n)



or

l = [n for n in range(1, 11)]



you should write only

l = range(1, 11)



In the two top code blocks you’re creating a new list, iterating through it and just returning each element. It’s just an expensive way of creating a list copy.

To get a new dictionary with all keys set to the same value based on another dict, do this:

old_dict = {'a': 1, 'c': 3, 'b': 2}
new_dict = { key:'your value here' for key in old_dict.keys()}



You’re receiving a SyntaxError because when you write

d = {}
d[i for i in range(1, 11)] = True



you’re basically saying: “Set my key ‘i for i in range(1, 11)’ to True” and “i for i in range(1, 11)” is not a valid key, it’s just a syntax error. If dicts supported lists as keys, you would do something like

d[[i for i in range(1, 11)]] = True



and not

d[i for i in range(1, 11)] = True



but lists are not hashable, so you can’t use them as dict keys.

d[tuple([i for i in range(1,11)])] = True