python 3.x - understanding behavior of mapping to an array -


when map modify array in place? know preferred way iterate on array list comprehension, i'm preparing algorithm ipyparallel, apparently uses map function. each row of array set of model inputs, , want use map, in parallel, run model each row. i'm using python 3.4.5 , numpy 1.11.1. need these versions compatibility other packages.

this simple example creates list , leaves input array intact, expected.

grid = np.arange(25).reshape(5,5) grid  array([[ 0,  1,  2,  3,  4],        [ 5,  6,  7,  8,  9],        [10, 11, 12, 13, 14],        [15, 16, 17, 18, 19],        [20, 21, 22, 23, 24]])  def f(g):     return g + 1  n = list(map(f, grid)) grid  array([[ 0,  1,  2,  3,  4],        [ 5,  6,  7,  8,  9],        [10, 11, 12, 13, 14],        [15, 16, 17, 18, 19],        [20, 21, 22, 23, 24]]) 

but when function modifies slice of input row, array modified in place. can explain behavior?

def f(g):     g[:2] = g[:2] + 1     return g  n = list(map(f, grid)) grid  array([[ 1,  2,  2,  3,  4],        [ 6,  7,  7,  8,  9],        [11, 12, 12, 13, 14],        [16, 17, 17, 18, 19],        [21, 22, 22, 23, 24]]) 


Comments

Popular posts from this blog

unity3d - Rotate an object to face an opposite direction -

angular - Is it possible to get native element for formControl? -

javascript - Why jQuery Select box change event is now working? -