Post algorithms options

List of available algorithms: BasicHistory - ConstraintsHistory - Correlations - GradientSensitivity - KMeans - ObjConstrHist - OptHistoryView - ParallelCoordinates - ParetoFront - QuadApprox - RadarChart - Robustness - SOM - ScatterPlotMatrix - VariableInfluence -

BasicHistory

Description

The BasicHistory post processing plots any of the constraint or objective functions w.r.t. optimization iterations or sampling snapshots.

The plot method requires the list of variable names to plot. It is possible either to save the plot, to show the plot or both.

Options

  • data_list, list(str) - list of variable names

  • extension, str - file extension

  • file_path, str - the base paths of the files to export

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

ConstraintsHistory

Description

The ConstraintsHistory post processing plots the constraints functions history in lines charts with violation indication by color on background.

The plot method requires the list of constraint names to plot. It is possible either to save the plot, to show the plot or both.

Options

  • constraints_list, list(str) - list of constraint names

  • extension, str - file extension

  • file_path, str - the base paths of the files to export

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

Correlations

Description

The Correlations post processing builds scatter plots of correlated variables among design variables, outputs functions and constraints

The plot method considers all variable correlations greater than 95%. An other level value, a sublist of variable names or both can be passed as options. The x- and y- figure sizes can also be modified in option. It is possible either to save the plot, to show the plot or both.

Options

  • coeff_limit, bool - if the correlation between the variables is lower than coeff_limit, the plot is not made

  • extension, str - file extension

  • file_path, str - the base paths of the files to export

  • n_plots_x, int - number of horizontal plots

  • n_plots_y, int - number of vertical plots

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

GradientSensitivity

Description

The GradientSensitivity post processing builds histograms of derivatives of objective and constraints

The plot method considers the derivatives at the last iteration. The iteration can be changed in option. The x- and y- figure sizes can also be modified in option. It is possible either to save the plot, to show the plot or both.

Options

  • extension, str - file extension

  • figsize_x, int - size of figure in horizontal direction (inches)

  • figsize_y, int - size of figure in vertical direction (inches)

  • file_path, str - the base paths of the files to export

  • iteration, int - the iteration to plot sensitivities, if negative, use optimum

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

KMeans

Description

The KMeans post processing performs a k-means clustering on optimization history.

The default number of clusters is 5 and can be modified in option.

The k-means construction depends on the MiniBatchKMeans class of the cluster module of the scikit-learn library .

Options

  • n_clusters, Unknown - prescribed number of clusters

ObjConstrHist

Description

The ObjConstrHist post processing plots the constraint functions history in lines charts.

By default, all constraints are considered. A sublist of constraints can be passed as options. It is possible either to save the plot, to show the plot or both.

Options

  • constr_names, list(str) - names of the constraints to plot

  • extension, str - file extension

  • file_path, str - the base paths of the files to export

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

OptHistoryView

Description

The OptHistoryView post processing performs separated plots: the design variables history, the objective function history, the history of hessian approximation of the objective, the inequality constraint history, the equality constraint history, and constraints histories.

By default, all design variables are considered. A sublist of design variables can be passed as options. Minimum and maximum values for the plot can be passed as options. The objective function can also be represented in terms of difference w.r.t. the initial value It is possible either to save the plot, to show the plot or both.

Options

  • extension, str - file extension

  • file_path, str - the base paths of the files to export

  • obj_max, float - maximum value for the objective in the plot

  • obj_min, float - minimum value for the objective in the plot

  • obj_relative, bool - plot the objective value difference with the initial value

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

  • variables_names, list(str) - list of the names of the variables to display

ParallelCoordinates

Description

The ParallelCoordinates post processing builds parallel coordinates plots among design variables, outputs functions and constraints

x- and y- figure sizes can be changed in option. It is possible either to save the plot, to show the plot or both.

Options

  • extension, str - file extension

  • figsize_x, Unknown - X size of the figure Default value = 10

  • figsize_y, Unknown - Y size of the figure Default value = 2

  • file_path, str - the base paths of the files to export

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

ParetoFront

Description

Compute the Pareto front Search for all non dominated points, ie there exists j such that there is no lower value for obj_values[:,j] that does not degrade at least one other objective obj_values[:,i].

Generates a plot or a matrix of plots if there are more than 2 objectives. Plots in red the locally non dominated points for the currrent two objectives. Plot in green the globally (all objectives) Pareto optimal points.

Options

  • extension, str - file extension

  • figsize_x, int - size of figure in horizontal direction (inches)

  • figsize_y, int - size of figure in vertical direction (inches)

  • file_path, str - the base paths of the files to export

  • objectives, list(str) - the functions names or design variables to plot if None, use the objective function (may be a vector)

  • objectives_labels, Unknown -

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

QuadApprox

Description

The QuadApprox post processing performs a quadratic approximation of a given function from an optimization history and plot the results as cuts of the approximation.

The function index can be passed as option. It is possible either to save the plot, to show the plot or both.

Options

  • extension, str - file extension

  • file_path, str - the base paths of the files to export

  • func_index, int - functional index

  • function, str - function name to build quadratic approximation

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

RadarChart

Description

Plot on radar style chart a list of constraint functions.

This class has the responsability of plotting on radar style chart a list of constraint functions at a given iteration.

By default, the iteration is the last one. It is possible either to save the plot, to show the plot or both.

Options

  • constraints_list, list(str) - list of constraints names

  • extension, str - file extension

  • figsize_x, Unknown -

  • figsize_y, Unknown -

  • file_path, str - the base paths of the files to export

  • iteration, int - number of iteration to post process

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

Robustness

Description

The Robustness post processing performs a quadratic approximation from an optimization history, and plot the results as cuts of the approximation computes the quadratic approximations of all the output functions, propagate analytically a normal distribution centered on the optimal design variable with a standard deviation which is a percentage of the mean passed in option (default: 1%) and plot the corresponding output boxplot.

It is possible either to save the plot, to show the plot or both.

Options

  • extension, str - file extension

  • file_path, str - the base paths of the files to export

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

  • stddev, float - standard deviation of inputs as fraction of x bounds

SOM

Description

The SOM post processing perform a self organizing map clustering on optimization history

Options of the plot method are the figure width and height, and the x- and y- number of cells in the SOM. It is also possible either to save the plot, to show the plot or both.

Options

  • annotate, Unknown - add label of neuron value to SOM plot

  • extension, str - file extension

  • file_path, str - the base paths of the files to export

  • height, Unknown - figure height

  • n_x, int - x-size

  • n_y, int - y-size

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

  • width, Unknown - figure width

ScatterPlotMatrix

Description

The ScatterPlotMatrix post processing builds scatter plot matrix among design variables, outputs functions and constraints.

The list of variable names has to be passed as arguments of the plot method. x- and y- figure sizes can be changed in option. It is possible either to save the plot, to show the plot or both.

Options

  • extension, str - file extension

  • figsize_x, int - size of figure in horizontal direction (inches)

  • figsize_y, int - size of figure in vertical direction (inches)

  • file_path, str - the base paths of the files to export

  • save, bool - if True, exports plot to pdf

  • show, bool - if True, displays the plot windows

  • variables_list, list(str) - the functions names or design variables to plot

VariableInfluence

Description

The VariableInfluence post processing performs first order variable influence analysis by computing df/dxi * (xi* - xi0) where xi0 is the initial value of the variable and xi* is the optimal value of the variable

Options of the plot method are the x- and y- figure sizes, the quantile level, the use of a logarithmic scale and the possibility to save the influent variables indices as a numpy file It is also possible either to save the plot, to show the plot or both.

Options

  • absolute_value, bool - if true, plot the absolute value of the influence

  • extension, str - file extension

  • figsize_x, int - size of figure in horizontal direction (inches)

  • figsize_y, int - size of figure in vertical direction (inches)

  • file_path, str - the base paths of the files to export

  • log_scale, bool - if True, use a logarithmic scale

  • quantile, float - between 0 and 1, proportion of the total sensitivity to use as a threshold to filter the variables

  • save, bool - if True, exports plot to pdf

  • save_var_files, bool - save the influent variables indices as a numpy file

  • show, bool - if True, displays the plot windows