@@ -58,9 +58,14 @@ function run_benchmarks(problems, optimizers)
58
58
optz = length(optimizers)
59
59
n = length(problems)
60
60
61
+ @info "here 1"
62
+
61
63
broadcast(c -> sizehint!(c, optz * n), [problem, n_vars, secs, solver, retcode])
62
64
65
+ @info "here 2"
66
+
63
67
for prob_name in problems
68
+ @info prob_name
64
69
nlp_prob = CUTEstModel(prob_name)
65
70
prob = OptimizationNLPModels.OptimizationProblem(nlp_prob, Optimization.AutoForwardDiff())
66
71
for optimizer in optimizers
@@ -93,27 +98,34 @@ The following figure shows the results of the same benchmarks previously describ
93
98
problems on this section.
94
99
95
100
```julia
101
+ @info "before"
96
102
eq_bou_problems = CUTEst.select(min_con=1, only_equ_con=true, only_free_var=false)
103
+ @info "after1"
97
104
98
105
# Analysis
99
106
eq_bou_results = run_benchmarks(eq_bou_problems, optimizers)
107
+ @info "after2"
100
108
101
109
@df eq_bou_results scatter(:n_vars, :secs,
102
110
group = :solver,
103
111
xlabel = "n. variables",
104
112
ylabel = "secs.",
105
113
title = "Time to solution by optimizer and number of vars",
106
114
)
115
+ @info "after3"
107
116
```
108
117
109
118
Next, we examine the same relationship for problems with inequality-constrained problems,
110
119
of which there are 244.
111
120
112
121
```julia
122
+ @info "after4"
113
123
neq_bou_problems = CUTEst.select(min_con=1, only_ineq_con=true, only_free_var=false)
124
+ @info "after5"
114
125
115
126
# Analysis
116
127
neq_bou_results = run_benchmarks(neq_bou_problems, optimizers)
128
+ @info "after6"
117
129
118
130
@df neq_bou_results scatter(:n_vars, :secs,
119
131
group = :solver,
@@ -126,4 +138,4 @@ neq_bou_results = run_benchmarks(neq_bou_problems, optimizers)
126
138
```julia, echo = false
127
139
using SciMLBenchmarks
128
140
SciMLBenchmarks.bench_footer(WEAVE_ARGS[:folder],WEAVE_ARGS[:file])
129
- ```
141
+ ```
0 commit comments