Extracting multiple sets of rows/ columns from a 2D numpy array

Question

I have a 2D numpy array from which I want to extract multiple sets of rows/ columns.

# img is 2D array
img = np.arange(25).reshape(5,5)
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]])

I know the syntax to extract one set of row/ column. The following will extract the first 4 rows and the 3rd and 4th column as shown below

img[0:4, 2:4]
array([[ 2,  3],
       [ 7,  8],
       [12, 13],
       [17, 18]])

However, what is the syntax if I want to extract multiple sets of rows and/or columns? I tried the following but it leads to an invalid syntax error

img[[0,2:4],2] 

The output that I am looking for from the above command is

array([[ 2],
       [12],
       [17]])

I tried searching for this but it only leads to results for one set of rows/ columns or extracting discrete rows/ columns which I know how to do, like using np.ix.

For context, the 2D array that I am actually dealing with has the dimensions ~800X1200, and from this array I want to extract multiple ranges of rows and columns in one go. So something like img[[0:100, 120:210, 400, 500:600], [1:450, 500:550, 600, 700:950]].

Answer 1

You can create your slices with numpy.r_:

np.r_[0:2, 4]
# array([0,1,4])

Then you can get the specific rows and columns as follows:

rows = np.r_[0:2, 4]
cols = np.r_[0, 2:4]
img[rows][:, cols]
# array([[ 0,  2,  3],
#       [ 5,  7,  8],
#       [20, 22, 23]])

Answer 2

IIUC, you can use numpy.r_ to generate the indices from the slice:

img[np.r_[0,2:4][:,None],2] 

output:

array([[ 2],
       [12],
       [17]])

intermediates:

np.r_[0,2:4]
# array([0, 2, 3])

np.r_[0,2:4][:,None]  # variant: np.c_[np.r_[0,2:4]]
# array([[0],
#        [2],
#        [3]])