通过 executor.map 传递参数

Question

How can I send another argument via executor.map function? See an example of code:

from concurrent.futures import ProcessPoolExecutor
from dask.dataframe import read_csv


def apply(row_of_small_df):
    # Here there is no access to big_df
    return


def main():
    small_df = read_csv('...')
    big_df = read_csv('...')
    with ProcessPoolExecutor() as executor:
        results = executor.map(apply, small_df.iterrows())
        for result in results:
            pass


if __name__ == '__main__':
    main()

Answer 1

Another alternative is using functools.partial :

Return a new partial object which when called will behave like func called with the positional arguments args and keyword arguments keywords. If more arguments are supplied to the call, they are appended to args. If additional keyword arguments are supplied, they extend and override keywords.

from functools import partial

def apply(big_df, row_of_small_df):
    # requires big_df to be passed in
    return

def main():
    small_df = read_csv('...')
    big_df = read_csv('...')

    apply_with_big_df = partial(apply, big_df)

    with ProcessPoolExecutor() as executor:
        results = executor.map(apply_with_big_df, small_df.iterrows())
        for result in results:
            pass

Answer 2

use a lambda:

#...
def apply(big_df, row_of_small_df):
    pass
#...
results = executor.map(lambda row_of_small: apply(big_df, row_of_small), small_df.iterrows())
#...

Answer 3

Use the zip function like so

results = executor.map(apply, zip(big_df.iterrows(), small_df.iterrows()))

The function should now be

def apply(params):
    big, small = params
    # your code