Y may be any array.  X is a simple integer scalar or vector.  The axis specification is optional.  If present, K must be a simple integer scalar or 1-element vector.  The value of K must be an axis of Y.  If absent, the last axis of Y is implied.  The form R←X⍀Y implies the first axis.  If Y is a scalar, it is treated as a one-element vector.

If Y has length 1 along the K^th (or implied) axis, it is extended along that axis to match the number of positive elements in X. Otherwise, the number of positive elements in X must be the length of the K^th (or implied) axis of Y.

R is composed from the sub-arrays along the K^th axis of Y. If X[I] (an element of X) is the J^th positive element in X, then the J^th sub-array along the K^th axis of Y is replicated X[I] times. If X[I] is negative, then a sub-array of fill elements of Y is replicated |X[I] times and inserted in relative order along the K^th axis of the result. If X[I] is zero, it is treated as the value ¯1. The shape of R is the shape of Y except that the length of the K^th axis is +/1⌈|X.

      0\⍳0
0
 
      1 ¯2 3 ¯4 5\'A'
A  AAA    AAAAA
 
      M
1 2 3
4 5 6
 
      1 ¯2 2 0 1\M
1 0 0 2 2 0 3
4 0 0 5 5 0 6
 
      1 0 1⍀M
1 2 3
0 0 0
4 5 6
 
      1 0 1\[1]M
1 2 3
0 0 0
4 5 6
 
      1 ¯2 1\(1 2)(3 4 5)
 1 2  0 0  0 0  3 4 5