https://julialang.org/" rel="nofollow">https://julialang.org/ release Platform Info: OS: Linux (x86_64-linux-gnu) CPU: 8 × AMD Ryzen 3 5300U with Radeon Graphics WORD_SIZE: 64 LLVM: libLLVM-15.0.7 (ORCJIT, znver2) Threads: 11 on 8 virtual cores julia> using LinearAlgebra, DynamicPolynomials julia> function f(n) @PolyVar a b c d e diagm( -2 => fill(a, n - 2), -1 => fill(b, n - 1), 0 => fill(c, n), 2 => fill(e, n - 2), 1 => fill(d, n - 1), ) end f (generic function with 1 method) julia> const m15 = f(15); julia> const m16 = f(16); julia> @time det(m15); 1.945673 seconds (45.22 M allocations: 2.261 GiB, 20.60% gc time, 4.02% compilation time) julia> @time det(m15); 1.991062 seconds (45.22 M allocations: 2.261 GiB, 23.74% gc time) julia> @time det(m16); 4.596664 seconds (106.67 M allocations: 5.324 GiB, 22.65% gc time) julia> @time det(m16); 4.648503 seconds (106.67 M allocations: 5.324 GiB, 22.66% gc time) ``` The above REPL session is with this commit applied, and with all other recent PRs of mine applied, to MultivariatePolynomials.jl, DynamicPolynomials.jl, and MutableArithmetics.jl. The same computation with MultivariatePolynomials v0.5.3 ran for multiple minutes before I decided to just kill it. Depends on JuliaAlgebra#285. Fixes JuliaAlgebra#281.

Commits

Commits on Nov 27, 2023

LinearAlgebra.det improvements

`LinearAlgebra.det` improvements