optilab.plotting package

Submodules

optilab.plotting.box_plot module

Plotting optimization results with box plot.

plot_box_plot(data: Dict[str, List[float]], savepath: str | None = None, *, show: bool = True, function_name: str | None = None, hide_outliers: bool = False) → None[source]

Plot box plots of optimization results.

Parameters:

  • data – dictionary where keys are optimization method or function names and values are list of best values from each run.

  • savepath – Path to save the plot, optional.

  • show – Wheather to show the plot, default True.

  • function_name – Name of the optimized function, used in title.

  • hide_outliers – If true, outliers won’t be shown on the box plot. Defualt is False.

optilab.plotting.convergence_curve module

Calculating and plotting the convergence curve.

convergence_curve(log: PointList) → List[float][source]

For a given log return a convergence curve - the lowest value achieved so far.

Parameters:

log – Results log - the values of errors of the optimized function.

Returns:

y values of the convergence curve.

plot_convergence_curve(data: Dict[str, List[PointList]], savepath: str | None = None, *, show: bool = True, function_name: str | None = None) → None[source]

Plot the convergence curves of a few methods using pyplot.

Parameters:

  • data – Lists of error logs of a few methods expressed as {method name: [log]}.

  • savepath – Path to save the plot, optional.

  • show – Wheather to show the plot, default True.

  • function_name – Name of the optimized function, used in title.

optilab.plotting.ecdf_curve module

Plotting and calculating ECDF curves.

_ecdf_thresholding(log: List[float], thresholds: List[float], n_dimensions: int, extend_to_len: int | None = None) → Tuple[List[float], List[float]][source]

Perform thresholding of a log with a given list of thresholds. The resulting y is the number of thresholds achieved by the log items.

Parameters:

  • log – Error log of a optimization function.

  • thresholds – ECDF value thresholds.

  • n_dimensions – Dimensionality of optimized function.

Returns:

x and y values for the curve.

ecdf_curve(data: Dict[str, List[PointList]], n_dimensions: int, allowed_error: float, n_thresholds: int = 100) → Dict[str, Tuple[List[float], List[float]]][source]

Calculate ECDF curves.

Parameters:

  • data – Lists of value logs indexed by method name.

  • n_dimensions – Dimensionality of the solved problem.

  • allowed_error – Tolerable error value, used as the last threshold.

  • n_thresholds – Number of ECDF thresholds.

Returns:

x, y plot points for each method.

plot_ecdf_curves(data: Dict[str, PointList], n_dimensions: int, allowed_error: float, n_thresholds: int = 100, savepath: str | None = None, *, show: bool = True, function_name: str | None = None) → None[source]

Calculate and plot ECDF curves.

Parameters:

  • data – Lists of value logs for every method.

  • n_dimensions – Dimensionality of the optimized function.

  • allowed_error – Tolerable error value, used as the last threshold.

  • n_thresholds – Number of ECDF thresholds.

  • savepath – Path to save the plot, optional.

  • show – Wheather to show the plot, default True.

  • function_name – Name of the optimized function, used in title.

Module contents

Module optilab.plotting . Contains functions to plot results.