function - python functools.partial with fixed array valued argument -

i'm trying vectorize following function on argument tiling:

def find_tile(x,tiling):     """     calculates index of closest element of 'tiling' 'x'.     tiling: array of grid positions     x: variable of same type elements of tiling     """     return np.argmin(np.linalg.norm(tiling - x, axis=1)) 

for instance, non-vectorized version of function can accept following arguments

tiling = np.array( [[i,j] in xrange(3) j in xrange(3)] ) x = np.array([1.2, 2.7]) 

i'm interested in finding fastest possible vectorisation, such x remains single vector , can pass list of arguments tiling

so tried defining multiple tilings using generator:

tilings = (tiling + np.random.uniform(0,1,2)  j in xrange(3)) 

and using map , functools.partial:

map(functools.partial(find_tile, x=x), tilings) 

apparently, there's problem x being array or something, since i'm getting error:

traceback (most recent call last):   file "main.py", line 43, in <module>     inds = map(functools.partial(find_tile, x=x), ts) typeerror: find_tile() got multiple values keyword argument 'x' 

can explain me how around it?

also, there alternative , faster way (possibly re-writing function find_tile?)

you passing in x keyword argument. map() passes in each element tilings in positional argument. however, since first positional argument x, clashes keyword argument. using name keyword argument not prevent same name being filled positional argument.

don't use keyword argument x; pass in positional argument partial():

map(functools.partial(find_tile, x), tilings) 

now each element tilings passed in second positional argument , call works.