Hstack, vstack and dstack in numpy

Recently python Code always sees np.hstack, np.vstack as well as np.dstack etc. , Today, I would like to briefly summarize my own understanding . These three stack The changes of can be called horizontal stack（hstack）, vertical stack（vstack） And depth stack（dstack）, The first two are well understood in two-dimensional arrays , But it is not easy to understand the three-dimensional situation .

1. Two dimensional array

Sample code ：

>>> a = np.arange(1,10)
>>> a = a.reshape(3,3)
>>> b = np.arange(10,19)
>>> b = b.reshape(3,3)
>>> np.hstack((a,b))
array([[ 1,  2,  3, 10, 11, 12],
       [ 4,  5,  6, 13, 14, 15],
       [ 7,  8,  9, 16, 17, 18]])
>>> np.vstack((a,b,c))
array([[ 1,  2,  3],
       [ 4,  5,  6],
       [ 7,  8,  9],
       [10, 11, 12],
       [13, 14, 15],
       [16, 17, 18]])
>>> np.dstack((a,b,c))
array([[[ 1, 10],
        [ 2, 11],
        [ 3, 12]],

       [[ 4, 13],
        [ 5, 14],
        [ 6, 15]],

       [[ 7, 16],
        [ 8, 17],
        [ 9, 18]]])

2. Three dimensional array

>>> a = np.arange(1,28)
>>> a = a.reshape(3,3,3)
>>> b = np.arange(28,55)
>>> b = b.reshape(3,3,3)

>>> np.hstack((a,b))
array([[[ 1,  2,  3],
        [ 4,  5,  6],
        [ 7,  8,  9],
        [28, 29, 30],
        [31, 32, 33],
        [34, 35, 36]],

       [[10, 11, 12],
        [13, 14, 15],
        [16, 17, 18],
        [37, 38, 39],
        [40, 41, 42],
        [43, 44, 45]],

       [[19, 20, 21],
        [22, 23, 24],
        [25, 26, 27],
        [46, 47, 48],
        [49, 50, 51],
        [52, 53, 54]]])

>>> np.vstack((a,b))
array([[[ 1,  2,  3],
        [ 4,  5,  6],
        [ 7,  8,  9]],

       [[10, 11, 12],
        [13, 14, 15],
        [16, 17, 18]],

       [[19, 20, 21],
        [22, 23, 24],
        [25, 26, 27]],

       [[28, 29, 30],
        [31, 32, 33],
        [34, 35, 36]],

       [[37, 38, 39],
        [40, 41, 42],
        [43, 44, 45]],

       [[46, 47, 48],
        [49, 50, 51],
        [52, 53, 54]]])

>>> np.dstack((a,b))
array([[[ 1,  2,  3, 28, 29, 30],
        [ 4,  5,  6, 31, 32, 33],
        [ 7,  8,  9, 34, 35, 36]],

       [[10, 11, 12, 37, 38, 39],
        [13, 14, 15, 40, 41, 42],
        [16, 17, 18, 43, 44, 45]],

       [[19, 20, 21, 46, 47, 48],
        [22, 23, 24, 49, 50, 51],
        [25, 26, 27, 52, 53, 54]]])

Conclusion ： Practical level , The vertical and depth are intuitive in a two-dimensional array , But the stack used to mark 3D arrays is very confusing , therefore vstack Recorded as on page 0 Dimension extension ,hstack Remember to expand in the first dimension ,dstack Recorded as on page 2 Dimension extension is enough . If you treat a three-dimensional array as an image , that ,hstack Indicates the horizontal splicing of two images ,vstack Splice the two images vertically ,dstack Superimpose the channels of the image .