numpy：二维数组，在第一个索引上运行时出错

Question

I'm facing a problem when trying to assign a specific value to a 100x100 2d array. I want it to be zero everywhere except along two segments: first line from the 25th to the 75th and last column from the 25th row to the 75th.

I tried doing it using the ": looping", I mean:

my_array[0][25:75] = 1.
my_array[25:75][99] = 1.

but the second line returns me the error:

IndexError: index 99 is out of bounds for axis 0 with size 50

As if it was first cutting an array of length 50 (75-25) and then assigning with it.

Of course I can get what I want using:

for x in xrange(25,75):
    my_array[x][-1] = 1.

But, out of curiosity, how could I get it done with : manipulation?

Answer 1

Put both array slices inside the same square brackets. Making your problem smaller so it's easier to show:

>>> my_array = np.zeros((10,10))
>>> my_array[0, 2:7] = 1
>>> my_array[2:7, -1] = 1
>>> my_array
array([[ 0.,  0.,  1.,  1.,  1.,  1.,  1.,  0.,  0.,  0.],
       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.],
       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.],
       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.],
       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.],
       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.],
       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.],
       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.],
       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.],
       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.]])

What's happening now is that you're requesting a 50-row slice of this array, and then the [99] is trying to get the 99th row of that , but there are only 50.