Metatheory.jl
Metatheory.jl copied to clipboard
Q: Metatheory.jl code optimization for multivariate polynomial expressions?
Hello,
After going through the Metatheory.jl documentation, there are a number of concepts that are novel and unfamilair to me, so I'm unable to discern if the package can do the following:
It is possible to use Metatheory.jl for code optimization in the sense of reducing the number of basic operations {+, -, x, /, ^} for large multivariate polynomial and polynomial-like expressions?
I'm currently using FORM 4.2.1 which performs a stochastic local search via stochastic hill climbing to find the near-minimal number of operations for a multivariate Horner scheme combined with CSEE [Common Subexpression Elimination]. FORM reduces the number of basic operations by about a factor of eight.
Is it possible to use Metatheory.jl to do the same or equivalent? It would be both useful and much simpler, from a coding perspective, to have an all-Julia solution.