Python: joblib does not work on custom-defined function

Question

I am trying to use parallel computation package joblib in python. I can execute the following example and get the result

Parallel(n_jobs=8)(delayed(sqrt)(i) for i in range(10))

However, the following code does not work out.

from joblib import Parallel, delayed
def f(x):
    return 1
def y(x):
    result=Parallel(n_jobs=8)(delayed(x)(i) for i in range(10))
    return result
if __name__ == '__main__':
    print y(f)

when I run the above code, it keeps running forever without generating any result or message.

Can anyone figure out the reason for this strange behavior(I am using windows)? thanks

Answer 1

The problem is in this line of code:

result=Parallel(n_jobs=8)(delayed(x)(i) for i in range(10))

It should instead be:

result=Parallel(n_jobs=8)(delayed(f)(i) for i in range(10))

You have to pass the function name to delayed, just like you did with the first example using the sqrt.