# FiniteSetAggregateFunctions2(S, A, R, B)¶

aggcat2.spad line 122 [edit on github]

S: SetCategory

A: FiniteSetAggregate S

R: SetCategory

B: FiniteSetAggregate R

`FiniteSetAggregateFunctions2`

provides functions involving two finite set aggregates where the underlying domains might be different. An example of this is to create a set of rational numbers by mapping a function across a set of integers, where the function divides each integer by 1000.

- map: (S -> R, A) -> B
`map(f, a)`

applies function`f`

to each member of aggregate`a`

, creating a new aggregate with a possibly different underlying domain.

- reduce: ((S, R) -> R, A, R) -> R
`reduce(f, a, r)`

applies function`f`

to each successive element of the aggregate`a`

and an accumulant initialised to`r`

. For example,`reduce(_+\$Integer, [1, 2, 3], 0)`

does a`3+(2+(1+0))`

. Note: third argument`r`

may be regarded as an identity element for the function.

- scan: ((S, R) -> R, A, R) -> B
`scan(f, a, r)`

successively applies`reduce(f, x, r)`

to more and more leading sub-aggregates`x`

of aggregate`a`

. More precisely, if`a`

is`[a1, a2, ...]`

, then`scan(f, a, r)`

returns spad {[reduce(f, [a1], r), reduce(f, [a1, a2], r), …]}.