I have a 5x5 matrix M and a function f that operates on vectors. r ← ⊃?⍴M is the index of a random row of M. I need to return a modified version of M without direct assignment (it's important for the larger program) where f is only applied to r⌷M. I'm doing this by mapping across M's rows, returning f⍵ if the row's index matches r, and just ⍵ if it doesn't. The function I came up with is:
({f(⍣(r = M⍳⍵))⍵}⍤1) M
It works, but it's not ideal. I don't like the r = M⍳⍵ part, because I'm searching for the index of ⍵ in every cycle. I think it'd make more sense to operate across ⍳5 instead, referencing each row in terms of each ⍵ in this vector. I can't seem to get this to work though.
Any help making my function less ugly is appreciated.
Example code:
M←5 5⍴⍳25
f←-
r←⊃?⍴M
⎕←f@r⊢M
Example result:
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