Hi All,
I wanted to return values of Range type so that it becomes clear in the end-users mind that we are talking about a number range and will be used as a pages iterator. But, I realize the only way to create a Range is with the function range. The Range object per-se is not documented and just mapped to UnionAll.
Is it the right usage or Range object? Or is it intended more from a set theoretic relationship like a domain range [1,55.5) use case. I see various mathematical and statistical operations on it.
I can return a Tuple{Int,Int} or Pair{Int, Int} but both do not provide the same clarity as Range.
regards,
Sambit
Range is an abstract type, so a value cannot have this type, only a subtype of Range.
For integers, UnitRange should be fine. You can use the a:b constructor. The most commonly understood meaning is “all integers between a and b, inclusive”, but you can use it for anything you like.
For floating point intervals, various packages have different implementations, with different semantics.
1 Like
Unfortunately, the Range object concrete types are not documented clearly. From the REPL help.
help?> UnitRange
search: UnitRange AbstractUnitRange
No documentation found.
Base.UnitRange is of type UnionAll.
Summary
≡≡≡≡≡≡≡≡≡
struct UnionAll <: Type{T}
Fields
≡≡≡≡≡≡≡≡
var :: TypeVar
body :: Any
Supertype Hierarchy
≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡
UnionAll <: Type{T} <: Any
help?> AbstractUnitRange
search: AbstractUnitRange
No documentation found.
Base.AbstractUnitRange is of type UnionAll.
Summary
≡≡≡≡≡≡≡≡≡
struct UnionAll <: Type{T}
Fields
≡≡≡≡≡≡≡≡
var :: TypeVar
body :: Any
Supertype Hierarchy
≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡
UnionAll <: Type{T} <: Any
I am sure pull requests for documentation would be welcome; however, most of the documentation for UnitRange would be found in the manual where : is discussed. These would be a good start.
1 Like