OptimizationPlotter#

Bases: object

Visualize the progress and results of an optimization.

Provides figures showing the objective values and input variables throughout an optimization, as well as the final population in input and objective space.

The plotter can be created from an optimization log or an optimizer instance. Each plotting method returns a Matplotlib Figure for further customization or saving.

Parameters:

datadict: Optimization results in the JSON log format produced by the optimizers.
param_nameslist of str, optional: Labels for the optimization variables. Defaults to param1, param2, etc.
fitness_nameslist of str, optional: Labels for the objectives. Defaults to fitness1, fitness2, etc.
exclude_failedbool, optional: If True, exclude evaluations assigned the non_execution_penalty from the plots (i.e. failed evaluations).
fitness_limitslist or dict, optional: Plotting limits for each objective, specified as (min, max) pairs. If dictionary is used, the keys should match the fitness names.
progress_seriessequence of str or str, optional: Curves to display in the optimization progress plot. Use "all" to display all available curves. The series applicable to all optimizations: (“best”, “candidates”). The series also applicable to evolution: (“min-max”, “min”, “max”, “median”, “mean”).
series_labelsdict, optional: Custom legend labels for progress_series curves, keyed by curve tag from the progress_series.
series_specsdict, optional: Custom plotting properties for individual curves, keyed by curve name. Each entry can specify kind and/or style. See examples below.
iteration_labelstr, optional: Label used for optimization iterations.
progress_titlestr, optional: Title of the optimization progress plot.
population_labelstr, optional: Label used for candidates in the final iteration.

Attributes:

runRunData: Parsed optimization results, including input and objective dimensions, labels, evaluation counts, and the run summary.

Examples

Create a plotter from a saved optimization log and customize the labels and title of the figures:

plotter = OptimizationPlotter.from_json(
    "Simulation_00001.json",
    param_names=["Well width (nm)", "Barrier width (nm)"],
    fitness_names=["Energy error (eV)"],
    progress_title="Quantum well optimization",
)

plotter.plot_all(show=True)

Customize the progress curves and objective limits for an evolutionary optimization. Include only candidates whose objective values are within the specified ranges, and customize the styles of the “best” and “mean” curves:

plotter = OptimizationPlotter.from_json(
    "Simulation_00001.json",
    fitness_names=["Energy error", "Efficiency"],
    fitness_limits={
        "Energy error": (0, 0.5),
        "Efficiency": (0, 1),
    },
    progress_series=["best", "mean", "min-max"],
    series_specs={
        "best": {"kind": "line", "style": {"linewidth": 2}},
        "mean": {"kind": "line", "style": {"linestyle": "--"}},
    },
)

plotter.plot_all(show=True)

classmethod from_json(json_path, **kwargs) → OptimizationPlotter#

Create a plotter from a saved optimization log.

Examples

Create a plotter from a saved optimization log and customize the labels.

plotter = OptimizationPlotter.from_json(
    "Simulation_00001.json",
    param_names=["Well width (nm)", "Barrier width (nm)"],
)

classmethod from_optimizer(optimizer, **kwargs) → OptimizationPlotter#

Create a plotter from a completed optimization.

Loads the optimization results recorded by the optimizer.

Parameters:

optimizerEvolution or LocalOptimizer or BayesianOptimization: An optimizer whose run has produced a log.
**kwargs: Additional plotting configuration passed to OptimizationPlotter.

Returns:

OptimizationPlotter: Plotter initialized with the optimizer’s results.

Examples

Create a plotter from a completed local optimization and customize the figure labels and title:

from nextnano_optimizers.optimizer import LocalOptimizer
from nextnano_optimizers.plotting import OptimizationPlotter

...

optimizer = LocalOptimizer(
    nextnanoio=io,
    metric=metric,
    bounds=[(3, 10), (2, 8)],
)

optimizer.run(method="Nelder-Mead")

plotter = OptimizationPlotter.from_optimizer(
    optimizer,
    param_names=["Well width (nm)", "Barrier width (nm)"],
    fitness_names=["Energy error (eV)"],
    progress_title="Quantum well optimization",
)

plotter.plot_all(show=True)

plot_all(save_dir=None, show=False, progress_series=None, fig_format='png')#

Generate all available optimization figures.

Generates the progress figure and, when applicable, figures showing the final population in input and objective space.

Parameters:

save_dirstr, default=None: Directory in which to save the figures. Created if it does not exist. If None, figures are not saved.
showbool, default=False: Display figures in a window.
progress_seriessequence of str or str, default=None: Curves to display in the progress figure. Overrides the plotter’s default progress_series configuration.
fig_formatstr, default=”png”: Format of the saved figures, such as "png", "pdf", or "svg".

Raises:

ValueError: If the requested figure format is not supported by Matplotlib.

plot_inputs()#

Plot the final population in input-variable space.

Shows the final population, with a separate scatter plot for each pair of input variables. Pareto-optimal candidates are highlighted.

Returns:

matplotlib.figure.Figure or None

plot_pareto()#

Plot the final population in objective space.

Shows the final population, with a separate scatter plot for each pair of objectives. Pareto-optimal candidates are highlighted.

For two objectives, the Pareto-optimal points are connected to illustrate the Pareto front.

Returns:

matplotlib.figure.Figure or None: Figure showing the final population in objective space, or None if fewer than two objectives are defined.

plot_progress(fig=None, series=None)#

Plot the progress of the optimization.

Shows the objective values and input variables of the evaluated candidates throughout the optimization. Each objective and input variable is displayed in a separate panel.

Parameters:

figmatplotlib.figure.Figure, optional

Existing figure to clear and reuse. If not provided, a new figure is created.

serieslist of str or str, optional

Progress curves to display. Available options are "candidates", "min-max", "min", "max", "median", "mean", and "best". Use "all" to display every curve.

Defaults to the curves configured through progress_series.

Returns:

matplotlib.figure.Figure: Figure showing the optimization progress.

Notes

The "best" curve shows the lowest objective value found so far and is displayed only for objectives.

If evaluations are excluded through exclude_failed or fitness_limits, an additional panel shows the number of excluded evaluations at each iteration.