Is there a structure that combines NammedArray and DotMap functionality?

General Usage
,
1 
x = NamedArray([1,2,3],["a","b","c"])
x = DotMap( Dict( zip(["a","b","c"],[1,2,3]) ) )

So you could write both exp.(x) and x.b

2

It wouldn’t be hard to make one, but why? this would determine the optimal underlying storage mechanism, or whether it’s even called for

3

Dictionaries or keyed arrays give you this functionality with almost the same syntax.

Dictionaries:

julia> using Dictionaries

julia> x = dictionary(zip(["a","b","c"], [1,2,3]))
3-element Dictionary{String, Int64}
 "a" │ 1
 "b" │ 2
 "c" │ 3

julia> exp.(x)
3-element Dictionary{String, Float64}
 "a" │ 2.718281828459045
 "b" │ 7.38905609893065
 "c" │ 20.085536923187668

julia> x["b"]
2

It’s an ordered dictionary where values can be mapped or broadcasted with the same performance as regular arrays.

And keyed arrays:

julia> using AxisKeys

julia> x = KeyedArray([1,2,3], ["a","b","c"])
1-dimensional KeyedArray(...) with keys:
↓   3-element Vector{String}
And data, 3-element Vector{Int64}:
 ("a")  1
 ("b")  2
 ("c")  3

julia> exp.(x)
1-dimensional KeyedArray(...) with keys:
↓   3-element Vector{String}
And data, 3-element Vector{Float64}:
 ("a")   2.718281828459045
 ("b")   7.38905609893065
 ("c")  20.085536923187668

julia> x("b")
2
4

Not a big thing.
x.b is just a bit more succinct than x["b"] or x("b")
so if you could use it then I would.

5

If you wanted to use KeyedArrays as above, it looks like this might work:

Base.getproperty(x::KeyedArray, k::Symbol) = x(string(k))

though you might want to check the source code to make sure it doesn’t break anything.

1 Like
6

Thank you. That’s a nice approach. If its dangerous I might just stick with a few extra characters x("b") :slight_smile: