Does the 'in' clause used on python dictionaries call the keys() function every time?

Question

Let's say I have a

dict = {...} #lots of words in dictionary  

I have to do a

for word in ...: #long list of words
    if word in dict:
        #do something

My question is, does the 'if word in dict' call the dict.keys() function every time and hence a lot slower than if I added another variable at the top that is dict_keys = dict.keys()? The result of what I am talking about would be this.

dict = {...}
dict_keys = dict.keys()

for word in ...:
    if word in dict_keys:
        #do something

Thanks

Answer 1

no. foo in mydict is in fact a lot faster than foo in keys_list, since dicts are hash tables, so finding the element inside it is O(1). While foo in keys_list would be O(n) (slower as the number of keys grows bigger)

But you could always test yourself:

$ python -m timeit -s "x = range(1000)" "15 in x"
1000000 loops, best of 3: 0.311 usec per loop
$ python -m timeit -s "x = dict.fromkeys(xrange(1000))" "15 in x"
10000000 loops, best of 3: 0.0515 usec per loop

So checking directly on the dict for 1000 elements is one order of magnitude faster, not considering the time of the .keys()