SolverBenchmark.jl

SolverBenchmark.jl copied to clipboard

Handle Bool in LaTeX

1

We have `supertype(Bool) = Integer`. Should there be a special case or 0-1?

abelsiqueira

Unicode characters

2

Unicode characters are not converted by `latex_table` function. It will be a nice feature to add automatic conversion for them. I wanted to use `✔` and `✗` for the status...

amontoison

Legend placement appears buggy

6

We set the legend to the `:bottomright` by default https://github.com/JuliaSmoothOptimizers/SolverBenchmark.jl/blob/master/src/profiles.jl#L66, but the result doesn't reflect that setting: https://github.com/JuliaSmoothOptimizers/JSOSolvers.jl/issues/2#issuecomment-529036660

dpo

Join with multicolumn

1

ref @dpo LaTeX allows multicolumn, so we can have joined solvers with that instead. Does PrettyTables allow it?

abelsiqueira

allow AbstractDict in profiles

1

This change should allow uses to use an OrderedDict to store statistics, and have solvers use consistent colors in plots when producing several profiles with variants of the same solvers.

dpo

Parallel benchmarks

1

Cleaned up the code and rebased on newest changes from https://github.com/JuliaSmoothOptimizers/SolverBenchmark.jl/pull/167

farhadrclass

Fix Examples CI

1

https://github.com/JuliaSmoothOptimizers/SolverBenchmark.jl/actions/runs/11805344788/job/32887581626?pr=167 The issue is that AmplNLReader does not support NLPModel linear API, see https://github.com/JuliaSmoothOptimizers/AmplNLReader.jl/issues/127 . So, using Percival on inequality constrained problem failed because it uses SlackModel that rely on...

tmigot

docs: document solver-specific statistics in solve_problems

2

Added a "Solver-specific statistics" section to the solve_problems docstring in `run_solver.jl`. The new text explains how entries in s.solver_specific are handled: - scalar (non-vector) entries become additional DataFrame columns; -...

arnavk23