Currently, given the output of an optimization:
out_opti = runExperimentOpti(...)
inspecting the parameters, gives vectors and its difficult to associate parameters.
julia> out_opti.parameters.optimized
37-element Vector{Any}:
0.7997645f0
0.17955744f0
0.36396587f0
⋮
0.07381531f0
Therefore, currently, I am constructing ComponentVector of the indices.
clean_symbol(s) = Symbol(replace(s , "." => "_"))
par_names = ("cTauSoilW.opt_soilW", "cTauSoilW.opt_soilW_c")
par_name = first(par_names)
par_ind = CA.ComponentVector(; map(par_names) do par_name
clean_symbol(par_name) => findfirst(==(par_name), out_opti.parameters.name_full)
end...)
out_opti.parameters.default[par_ind]
Suggestion: Provide some easier way to inspect the output.
Option1 : Provide outputs as ComponentVectors so that one can index by names
out_opti.parameters.optimized[:cTauSoilW_opt_soilW]
the default print method would also print the names together with the values.
Option 2: allow Key index (dot) with parameter names:
out_opti.parameters.cTauSoilW_opt_soilW.optimized
This is a low-priority issue. More a maybe-convenience.
Currently, given the output of an optimization:
out_opti = runExperimentOpti(...)inspecting the parameters, gives vectors and its difficult to associate parameters.
Therefore, currently, I am constructing ComponentVector of the indices.
Suggestion: Provide some easier way to inspect the output.
Option1 : Provide outputs as ComponentVectors so that one can index by names
out_opti.parameters.optimized[:cTauSoilW_opt_soilW]the default print method would also print the names together with the values.
Option 2: allow Key index (dot) with parameter names:
out_opti.parameters.cTauSoilW_opt_soilW.optimizedThis is a low-priority issue. More a maybe-convenience.