I'm working in python and currently have the following code:
list = []
for a in range(100):
for b in range(100):
for c in range(100):
list.append(run(a,b,c))
where run(a,b,c) returns an integer (for example, it could multiply the three numbers together). Is there a faster way to either loop over these numbers or use a map function?
Thanks :)
Have a look at the itertools-module and particulary the product method
example usage:
for i in itertools.product(range(0,100), repeat=3):
#do stuff with i
list.append(run(i[0],i[1],i[2]))
Note that the function call can be shortened to:
list.append(run(*i))
in the example above. see docs.python.org for explanation of Unpacking Argument Lists.
As an example, the output from product(range(0,2), repeat=3)) looks like this:
(0, 0, 0)
(0, 0, 1)
(0, 1, 0)
(0, 1, 1)
(1, 0, 0)
(1, 0, 1)
(1, 1, 0)
(1, 1, 1)
list.append(run(i[0],i[1],i[2])) makes my eyes hurt a lot more than my one line solution. The short version with run(*i) is acceptable, though.I think you can use imap to do this :
from itertools import imap
result = list(imap(run, range(100), range(100), range(100)))
imap yields its result... so if you want to iterate of the results don't use the list()
map in Python 3 behavior as imap in Python 2, and this code is equivalent to [ run(a, b, c) for a, b, c in zip(range(100), range(100), range(100)) ] instead of what you want.from itertools import product
my_list = [run(a, b, c) for a, b, c in product(xrange(100), xrange(100), xrange(100))]
Or:
from itertools import product
my_list = [run(a, b, c) for a, b, c in product(xrange(100), repeat=3)]
import itertools
li = [run(*triple) for triple in itertools.product(xrange(100), repeat=3)]
an alternative, depending on what you want to do exactly:
my_array = numpy.tile(numpy.arange(100),(3,1)).T
def run(row):
return numpy.prod(row, 0)
[run(row) for row in my_array]
Of course, it all depends on run, e.g. if you want to take the product, you could also operate on the whole array, which is much faster:
my_product = run(my_array.T)
my_product = numpy.prod(my_array, 1)
