f must be a dyadic function. X must be a simple scalar or one-item integer array. Y may be any array whose sub-arrays along the Kth axis are appropriate to function f.
The axis specification is optional. If present, K must identify an axis of Y. If absent, the last axis of Y is implied. The form R←Xf⌿Y implies the first axis of Y.
R is an array formed by applying function f between items of sub-vectors of length X taken from vectors along the Kth (or implied) axis of Y.
X can be thought of as the width of a "window" which moves along vectors drawn from the Kth axis of Y.
If X is zero, the result is a (⍴Y)+(-⍴⍴Y)↑1 array of identity elements for the function f. See Identity Elements.
If X is negative, each sub-vector is reversed before being reduced.
⍳4 1 2 3 4 3+/⍳4⍝ (1+2+3) (2+3+4) 6 9 2+/⍳4⍝ (1+2) (2+3) (3+4) 3 5 7 1+/⍳4⍝ (1) (2) (3) (4) 1 2 3 4 0+/⍳4⍝ Identity element for + 0 0 0 0 0 0×/⍳4⍝ Identity element for × 1 1 1 1 1 2,/⍳4⍝ (1,2) (2,3) (3,4) 1 2 2 3 3 4 ¯2,/⍳4⍝ (2,1) (3,2) (4,3) 2 1 3 2 4 3