CODEWITHSUNDEEP
typesjulia
Sergey
Sergey

Reputation: 1328

julia get value of type parameters

I have a parametric type. For example:

> Array([1 2;3 4])

Its type is

> typeof(Array([1 2;3 4]))

Array{Int64,2}

I can get the first type parameter using eltype:

> eltype(typeof(Array([1 2;3 4])))

Int64

How can I access second and possibly other type parameters?

Upvotes: 5

Views: 1182

Answers (2)

François Févotte
François Févotte

Reputation: 20248

If you're speaking specifically of (Abstract)Array types, then the dimension can be retrieved using ndims:

julia> ndims(Array{Int64, 2})
2


If, on the other hand, you want to write a custom function that extracts a given parameter from a parametric type, you can use build one like this:

julia> second_param(::Type{Array{T, N}}) where {T, N} = N
second_param (generic function with 1 method)

julia> second_param(Array{Int64, 2})
2

(I'm using Array here for the sake of the example, but the same kind of construct could be used to extract parameters from any other parametric type)

Upvotes: 6

Bogumił Kamiński
Bogumił Kamiński

Reputation: 69839

What François Févotte recommends is best and safest. However, if you wanted to dig into the internals (again - I would not recommend it in production code, but sometimes it is useful) then you can write:

get_parameters(x::DataType) = collect(x.parameters)

Now you can get a vector of parameters of x for any type that is DataType:

julia> get_parameters(Vector{Int})
2-element Array{Any,1}:
  Int64
 1

julia> get_parameters(Int)
0-element Array{Any,1}

julia> get_parameters(Dict{Int, Union{String, Missing}})
2-element Array{Any,1}:
 Int64
 Union{Missing, String}

The benefit here is that x you pass can be any DataType.

Upvotes: 3

Related Questions