Currently, I note gmse.R and gmse_apply.R each separately construct the paras vector, even though its structure, I think, is expected to be exactly the same in both; consisting partially of dynamic variables and partially of a set of constants referring to locations in arrays, etc.
I wonder whether it would be useful to have a separate function, whose only job it is to construct paras from only the set of variables it includes - which can be called from either gmse or gmse_apply?
I'd be happy to suggest a change that does this, but there may be limitations/constraints on this that I haven't identified yet.
Possible upside - it would avoid some replication in code between gmse.R and gmse_apply.R.
Possible downside - it would require a separate function (and possibly file) that can be sourced by both the above functions.
Currently, I note
gmse.Randgmse_apply.Reach separately construct theparasvector, even though its structure, I think, is expected to be exactly the same in both; consisting partially of dynamic variables and partially of a set of constants referring to locations in arrays, etc.I wonder whether it would be useful to have a separate function, whose only job it is to construct
parasfrom only the set of variables it includes - which can be called from eithergmseorgmse_apply?I'd be happy to suggest a change that does this, but there may be limitations/constraints on this that I haven't identified yet.
Possible upside - it would avoid some replication in code between
gmse.Randgmse_apply.R.Possible downside - it would require a separate function (and possibly file) that can be sourced by both the above functions.