1. Numpy traversal element processing function: np.apply_along_axis(np.mean(),axis=0,arr=a)
The first parameter is the operation function, which can be a custom function, the second parameter axis=0 is column-wise operation, 1 is row-wise operation, and the third parameter is the name of the operation array
2. Change the shape of the array a.reshape(3,4)
3. Array transpose transpose()
4. Array combination vertical insertion np.vstack(a,b) horizontal insertion np.hstack(a,b) multiple array combination column_stock(a,b,c) row_stock(a,b,c)
5. Array segmentation Horizontal segmentation hsplit(a,2) Vertical segmentation vsplit(a,2) Divide the array into two parts, unequal segmentation split(a,[1,5,10],axis=1 ); where the first parameter is an array, the second parameter is the split position, [1,5,10] means split into four pieces, 1 is the first piece, 2:5 is the second piece..., the first The three parameters 1 is to split by column, 0 is to split by row, that is, 1 column is the first block, 2:5 is the second block...