You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Agents: read the guide for agents in #310 first (order, difficulty, how to claim an issue, working method). The specification is in this description and in the comments below.
Move plot functions from CTBenchmarks into OCP. Can be useful to plot JuMP solutions.
Once [Paper] Port to OptimalControl v2 and new problem interface #283 builds a CTModels.Solution from a JuMP model (when the problem exists in OptimalControl), plotting both together is plot(sol_oc) then plot!(sol_jump), with CTModels' Plots extension. No plotting code moves to this package.
Of CTBenchmarks' src/plot_solutions.jl: the separate OptimalControl / JuMP plotting code becomes useless; the presentation styles (colour and marker per modeler × solver, layout per problem) stay in CTBenchmarks, on top of plot!(sol); the costate sign flip for JuMP (costate_multiplier) belongs to the Solution construction ([Paper] Port to OptimalControl v2 and new problem interface #283, [General] Costate #170), not to plotting.
Remaining here: check and document the overlay of OptimalControl and JuMP solutions (including N vs N+1 controls per scheme, #284); clean up CTBenchmarks when it is worked on.
The prerequisite is done: OptimalControlProblems.solution(model, pb) builds a CTModels.Solution from a solved JuMP model (#329, PR #353), so plot(sol_oc) then plot!(sol_jump) works for every problem available in OptimalControl. Problems available in JuMP only still wait for control-toolbox/CTModels.jl#515. Costates of JuMP have the opposite sign today (#326): plot them together only after #326.
Activity
ocots commentedon Nov 4, 2026
Scope reduced (decided in #310):
CTModels.Solutionfrom a JuMP model (when the problem exists in OptimalControl), plotting both together isplot(sol_oc)thenplot!(sol_jump), with CTModels' Plots extension. No plotting code moves to this package.src/plot_solutions.jl: the separate OptimalControl / JuMP plotting code becomes useless; the presentation styles (colour and marker per modeler × solver, layout per problem) stay in CTBenchmarks, on top ofplot!(sol); the costate sign flip for JuMP (costate_multiplier) belongs to theSolutionconstruction ([Paper] Port to OptimalControl v2 and new problem interface #283, [General] Costate #170), not to plotting.Solutiononce [Feature] build_solution from model information, without a full continuous Model CTModels.jl#515 is done.Remaining here: check and document the overlay of OptimalControl and JuMP solutions (including
NvsN+1controls per scheme, #284); clean up CTBenchmarks when it is worked on.ocots commentedon Nov 6, 2026
Update
The prerequisite is done:
OptimalControlProblems.solution(model, pb)builds aCTModels.Solutionfrom a solved JuMP model (#329, PR #353), soplot(sol_oc)thenplot!(sol_jump)works for every problem available in OptimalControl. Problems available in JuMP only still wait for control-toolbox/CTModels.jl#515. Costates of JuMP have the opposite sign today (#326): plot them together only after #326.🤖 Generated with Claude Code