def make_plots(benchmark_result: BenchmarkResult, problem_name: str,
include_individual: bool) -> List[AxPlotConfig]:
plots: List[AxPlotConfig] = []
# Plot objective at true best
plots.append(
optimization_trace_all_methods(
y_dict=benchmark_result.objective_at_true_best,
optimum=benchmark_result.optimum,
title=f"{problem_name}: cumulative best objective",
ylabel="Objective at best-feasible point observed so far",
))
if include_individual:
# Plot individual plots of a single method on a single problem.
for m, y in benchmark_result.objective_at_true_best.items():
plots.append(
optimization_trace_single_method(
y=y,
optimum=benchmark_result.optimum,
# model_transitions=benchmark_result.model_transitions[m],
title=f"{problem_name}, {m}: cumulative best objective",
ylabel="Objective at best-feasible point observed so far",
))
# Plot time
plots.append(
optimization_times(
fit_times=benchmark_result.fit_times,
gen_times=benchmark_result.gen_times,
title=f"{problem_name}: optimization times",
))
return plots
def make_plots(
benchmark_result: BenchmarkResult, problem_name: str, include_individual: bool
) -> List[AxPlotConfig]:
plots: List[AxPlotConfig] = []
# Plot objective at true best
ylabel = (
"Feasible Hypervolume"
if benchmark_result.is_multi_objective
else "Objective at best-feasible point observed so far"
)
plots.append(
optimization_trace_all_methods(
y_dict=benchmark_result.true_performance,
optimum=benchmark_result.optimum,
title=f"{problem_name}: Optimization Performance",
ylabel=ylabel,
)
)
if include_individual:
# Plot individual plots of a single method on a single problem.
for m, y in benchmark_result.true_performance.items():
plots.append(
optimization_trace_single_method(
y=y,
optimum=benchmark_result.optimum,
# model_transitions=benchmark_result.model_transitions[m],
title=f"{problem_name}, {m}: cumulative best objective",
ylabel=ylabel,
)
)
# Plot time
plots.append(
optimization_times(
fit_times=benchmark_result.fit_times,
gen_times=benchmark_result.gen_times,
title=f"{problem_name}: cumulative optimization times",
)
)
if benchmark_result.pareto_frontiers is not None:
plots.append(
plot_multiple_pareto_frontiers(
frontiers=not_none(benchmark_result.pareto_frontiers),
CI_level=0.0,
)
)
return plots