high dimensional matrix index

Community Support
,
1

https://learninghub.kx.com/forums/topic/high-dimensional-matrix-index

apologies if this is obvious for  Q professionals, I'm struggling to understand why running below in console returns 1 2 3 4 5 6,  not 5 6  ?  even ChatGPT says it returns 5 6...


mm:((1 2;3 4;5 6);(10 20;30 40;50 60))

mm 0 2

2

When indexing at depth you need the square brackets or index-at for nested indexing dot-apply

mm:((1 2;3 4;5 6);(10 20;30 40;50 60))
mm 0 2
1 2 3 4 5 6

0N!mm 0 2

((1 2;3 4;5 6);(long$();long$();`long$()))

1 2 3 4 5 6


mm[0;2]

5 6

mm . (0;2)

5 6

mm . 0 2


5 6

If you really wanted you could use simple apply and the iterator over to achieve what you are looking for

mm@/0 2

5 6

If you use 0N! you can see that when using mm 0 2 you’re basically indexing at the first level, first for index 0 which returns the first top level, then at index 2 which doesn’t exist hence you get a list of empty long lists
3. Lists - Q for Mortals
Apply (At), Index (At), Trap (At) | Reference | kdb+ and q documentation - kdb+ and q documentation



3

To return 5 6 you will need to index at depth by passing a semicolon

mm[0;2]
5 6

Indexing by passing a list of integers returns all indexes so in this case:

mm 0 // returns the element at position 0
1 2
3 4
5 6
mm 2 // returns the element at position 2 - which doesn't exist
    
mm 0 2 // get both elements at positions 0 and 2 - because none at 2 we just get the 0 elements back 
1 2 3 4 5 6

This concept covered in the free academy course in the module on List Amendment and eliding at depth.

https://learninghub.kx.com/courses/kdb-developer-level-2/lessons/lists/topic/list-indexing-access/


4

Thanks!  good point to know!

5

Thanks! maybe time to refresh my memory of the course did some time ago...