Using LinearSolve or A\b in @pre?

How-to
1

Hi,

I am using Pumas for Academia and I would like to do either

...
@pre 
...
A = ... #some matrix composed of model parameters
b = ... #some vector composed of model parameters

x = A\b
...

or

...
@pre 
...
A = ... #some matrix composed of model parameters
b = ... #some vector composed of model parameters

prob = LinearProblem(A, b)
x = solve(prob)
...

but both throw this stacktrace (quite long), which is I presume due to it trying to do both symbolically and some relevant operators not being defined. Is there a way to do this currently? Is there a way around the error?

Many thanks!

2

Upon some further research the error I am encountering seems to be resolved in Resolve Symbolics.Num - ForwardDiff.Dual ambiguities by asterycs · Pull Request #1036 · JuliaSymbolics/Symbolics.jl · GitHub but I assume it will take quite some time for that to trickle down to the academic Pumas version.