CODEWITHSUNDEEP
pythonsortingnumpy
Dan
Dan

Reputation: 403

NumPy Secondary Sorting

I'm having the following simple array : 

a = np.array([
             [1, 1000, 58],
             [1, 200, 69],
             [3, 300, 34],
             [4, 400, 82]], dtype = int)

When I issue the following lexsort() command : i = np.lexsort((a[:,0], a[:,1]))

I'm getting : 

array([[   1,  200,   69],
       [   3,  300,   34],
       [   4,  400,   82],
       [   1, 1000,   58]])

While I was expecting : 

array([[   1,  200,   69],
       [   1, 1000,   58],
       [   3,  300,   34],
       [   4,  400,   82],
   ])

Can anyone explain me why ? and how do I get the expected output ?

Upvotes: 0

Views: 83

Answers (1)

Dan
Dan

Reputation: 403

Following @Divakar comment. All I had to do is :

i = np.lexsort((a[:,1], a[:,0]))

Thank you

Upvotes: 3

Related Questions