Tips for passing Optional Arguments through the Julia Splatt Operator
Overview
In Julia, the most frequently used purpose of the ... splat is explained as the method of passing optional arguments. Basically, it uses the method of applying the splat operator to the tuple after determining in advance what options to put into which arguments, in the form of a named tuple.
Code
Passing to Multiple Functions
args1 = (; dims = 1)
The named tuple args1 above can be commonly used for all functions having an optional argument called dims. In the following example, sum() and minimum() are completely different functions, but they were applied regardless of the type of function as they both have dims.
julia> sum(rand(100,100); args1...)
1×100 Matrix{Float64}:
47.0704 45.7637 44.4513 48.2325 50.5745 51.9176 … 49.9548 47.6825 50.7284 50.0861 50.0168 50.5116
julia> minimum(rand(100,100); args1...)
1×100 Matrix{Float64}:
0.00702003 0.0163299 0.00665818 0.0174564 0.00589048 … 0.002967 0.00460205 0.0116248 0.0114521 0.0698425
Passing Multiple Arguments
args2 = (; dims = 2, rev = true)
The named tuple args2 above includes the optional argument rev in addition to dims. In the following example, the sort() function reflected both options well and returned the computation result regardless of the input data.
julia> sort(rand(0:9, 3,3); args2...)
3×3 Matrix{Int64}:
9 4 4
6 5 2
8 0 0
julia> sort(rand(3,3); args2...)
3×3 Matrix{Float64}:
0.438682 0.211154 0.108741
0.72113 0.445214 0.00910109
0.971441 0.666732 0.0227372
Entire Code
args1 = (; dims = 1)
sum(rand(100,100); args1...)
minimum(rand(100,100); args1...)
args2 = (; dims = 2, rev = true)
sort(rand(0:9, 3,3); args2...)
sort(rand(3,3); args2...)
Environment
- OS: Windows
- julia: v1.8.3