Question

UPDATE: Sorry I was wrong with my assumption of the problem. Its due to me trying to make the function parallel and have multiple proceses change the same dictionary. How can I design my function so it allows for this? If I have to send the dictionary(in this case) then it’ll keep overwriting itself and can’t span over all the cores. I’m happy to redesign the function but not sure how i can do it and still spread it over multiple cores.

To simplify, I basically have something like the following:

list = [1,2,3,10]
dictionary = {}
for x in list:
    for xx in range(100):
        for xxx in range(100):
            dictionary[x]=xx*xxx

This works, but when I wrap the for loop in a function, I get:
NameError: global name ‘dictionary’ is not defined

def test(x):
        for xx in range(100):
            for xxx in range(100):
                dictionary[x]=xx*xxx

I know this has something to do with the namespace but I don’t want to send variables to the function in this case because I think it’ll overwrite the variable dictionary. How can I reference the dictionary variable while in the function? The reason I ask is I want to create a process that spans multiple cpu’s, so I don’t want to declare the variables in each function.

I’m using functions but if I need to use classes or there’s another way to get around this then please me know and I’ll rethink my approach.

Answer 1

Try again:

lst = [1,2,3,10]
dictionary = {}
def test(x):
    for xx in range(100):
        for xxx in range(100):
            dictionary[x]=xx*xxx

for x in lst:
    test(x) 

You only get a NameError when a name is not defined. It has nothing to do with scopes.