SolverBenchmark.jl

SolverBenchmark.jl copied to clipboard

I think there are two possibilities for this. i) Update the file https://github.com/JuliaSmoothOptimizers/JSOTutorials.jl/blob/main/tutorials/introduction-to-solverbenchmark/index.jmd to add some comments on how the package handles `solver_specific` in the stats. See #149 for an...

tmigot

export profile data

16

d-monnet

`hdr_override` in `join` doesn't affect `invariant_cols`.

2

``` hdr_override = Dict(:name => "Name", :f => "f(x)", :t => "Time") df = join(stats, [:f, :t], invariant_cols=[:name], hdr_override=hdr_override) markdown_table(stdout, df) ``` ``` | id | name | f(x)_alpha |...

abelsiqueira

`performance_profile()` doesn't passe kwargs to `BenchmarkProfiles.performance_profiles()`

1

There is no option to pass `BenchmarkProfiles.performance_profiles()` keyword arguments such as `logscale` to `SolverBenchmark.performance_profiles`. As such it is not possible to desactivate the log scale (among other things) with `SolverBenchmark.performance_profiles`....

d-monnet

The goal of this PR is to allow the user to parallelize the evaluation of solvers when using the `bmark_solvers(solvers, problems;...)` method. It can be run in parallel locally or...

MonssafToukal

export profiles data

2

Hi there, I'd like to know if there is a function to export the data (.csv or similar) of the performance profiles (the coordinates of the points of the curves)....

d-monnet

dpo

When using `profile_solvers` for several plots, the figures are plotted horizontally. I wonder if there is a simple way to plot them vertically instead? I managed to make it work...

tmigot

Error in `pretty_stats` with `Vector` in the `GenericExecutionStats` specifics

2

``` using CUTEst, NLPModels, NLPModelsIpopt, SolverBenchmark, SolverCore #This package using DCISolver nmax = 100 _pnames = CUTEst.select( max_var = nmax, min_con = 1, max_con = nmax, only_free_var = true, only_equ_con...

tmigot

There's a confusion between the row number and the id field.

dpo