cocopp.genericsettings

module documentation

This module contains some global variables settings for COCO.

These variables may be modified (in particular in rungeneric1, -2, and -many) and are used for producing figures and tables.

For setting variables dynamically or changing other module settings here see config.py.

Variable appended_evals_minimal_trials minimum number of instances required in the appended_evals array such that DataSet.appended_evals is created to be different from DataSet.evals
Variable background algorithms to shown for comparison "in the background". Usage: `` background = {(color, linestyle): [path_to_data1, path_to_data2, ...], }`` Examples: background = {None: cocopp.archives.bbob.get_all('2009/')}...
Variable background_default_style Undocumented
Variable balance_instances give even weight to all instances by added columns with copies of underrepresented instances in DataSet.evals
Variable dim_related_colors Undocumented
Variable dim_related_markers Undocumented
Variable extraction_folder_prefix Undocumented
Variable figsize Undocumented
Variable figure_file_formats Undocumented
Variable filter_data_condition a callable which is used in rungenericmany to filter the list of DataSet. For example, to keep only the "effectively unimodal" functions of the 'bbob' testbed, set filter_data_condition = lambda ds: ds.funcId in [1, 2,] + list(range(5, 15))...
Variable force_assertions Undocumented
Variable foreground_algorithm_list a list of data files/folders as those specified in cocopp.main
Variable in_a_hurry Undocumented
Variable index_html_file_name Undocumented
Variable inputCrE Undocumented
Variable inputsettings Undocumented
Variable instancesOfInterest Undocumented
Variable instancesOfInterest2009 Undocumented
Variable instancesOfInterest2010 Undocumented
Variable instancesOfInterest2012 Undocumented
Variable instancesOfInterest2013 Undocumented
Variable instancesOfInterest2015 Undocumented
Variable instancesOfInterest2016 Undocumented
Variable instancesOfInterest2017 Undocumented
Variable instancesOfInterest2018 Undocumented
Variable instancesOfInterest2021 Undocumented
Variable instancesOfInterest2023 Undocumented
Variable instancesOfInterestBiobj2016 Undocumented
Variable instancesOfInterestBiobj2017 Undocumented
Variable interactive_mode Undocumented
Variable isConv Undocumented
Variable isExpensive Undocumented
Variable isFig Undocumented
Variable isLogLoss Undocumented
Variable isNoiseFree Undocumented
Variable isNoisy Undocumented
Variable isPickled Undocumented
Variable isRLDistr Undocumented
Variable isRldOnSingleFcts Undocumented
Variable isScatter Undocumented
Variable isTab Undocumented
Variable latex_commands_for_html Undocumented
Variable len_of_names_in_pprldmany_legend set the length, for example when we want to remove characters that are not fully displayed, 9 == len('best 2009')
Variable line_style_mapping map the input argument position to a line style position, by default the identity. This could be useful to get the same style for several algorithm variants. Only effective with when bool(parameter_sweep) is True...
Variable line_styles Undocumented
Variable many_algorithm_file_name Undocumented
Variable marker_size_multiplier multiply all marker sizes in below line_styles, used in compall.pprldmany.main
Variable maxevals_fix_display Undocumented
Variable minmax_algorithm_fontsize Undocumented
Variable minor_grid_alpha_in_pprldmany used in pprldmany for empirical runtime distributions, 0 means no minor grid
Variable outputdir Undocumented
Variable parameter_sweep configure line styles for a parameter sweep, see also cocopp.config.config_line_styles
Variable parameter_sweep_colormaps colormap(s) for parameter sweep set from the respective input long_option.
Variable parameter_sweep_sort assign color in sweep by leading float of algorithm name
Variable ppconv_file_name Undocumented
Variable ppfigcons1_file_name Undocumented
Variable ppfigcons_file_name Undocumented
Variable ppfigdim_file_name Undocumented
Variable ppfigs_file_name Undocumented
Variable pprldistr2_file_name Undocumented
Variable pprldmany_file_name Undocumented
Variable pprldmany_group_file_name Undocumented
Variable ppscatter_file_name Undocumented
Variable pptables_file_name Undocumented
Variable rcaxes Undocumented
Variable rcaxeslarger Undocumented
Variable rcfont Undocumented
Variable rcfontlarger Undocumented
Variable rclegend Undocumented
Variable rclegendlarger Undocumented
Variable rctick Undocumented
Variable rcticklarger Undocumented
Variable reference_algorithm_styles Undocumented
Variable runlength_based_targets Undocumented
Variable scaling_figures_with_boxes Undocumented
Variable scaling_plots_with_axis_labels Undocumented
Variable sequential_colormaps default color maps for a parameter sweep
Variable simulated_runlength_bootstrap_sample_size bootstrap samples, 30 is a multiple of 10 and 15. Used for tables and plots. int(1e4) would be preferable for a final camera-ready paper version.
Variable single_algorithm_file_name Undocumented
Variable single_runlength_factors Undocumented
Variable target_runlength Undocumented
Variable target_runlengths_in_scaling_figs Undocumented
Variable target_runlengths_in_single_rldistr Undocumented
Variable target_runlengths_pprldmany Undocumented
Variable target_runlengths_ppscatter Undocumented
Variable test Undocumented
Variable use_recommendations use only recommendations data (.mdat files) for the respective algorithm where the last element is recycled for the remaining algorithms
Variable verbose Undocumented
Variable warning_level Undocumented
Variable weight_evaluations_constraints weights used to sum function evaluations and constraints evaluations in the attribute DataSet._evals when data are loaded.
Variable xlimit_expensive maximal run length multiplier in expensive setting, used in config for pprldmany and ppfigdim
Variable xlimit_pprldmany maximal run length multiplier used in pprldmany.
Variable _current_args arguments found in rungeneric.main, namely a list of folders returned by COCODataArchive.get_extended of cocopp.archives.all.
Variable _default_line_styles Undocumented
appended_evals_minimal_trials: int =

minimum number of instances required in the appended_evals array such that DataSet.appended_evals is created to be different from DataSet.evals

background: dict =

algorithms to shown for comparison "in the background". Usage: `` background = {(color, linestyle): [path_to_data1, path_to_data2, ...], }`` Examples: background = {None: cocopp.archives.bbob.get_all('2009/')} for all algorithms benchmarked in 2009 on the bbob test suite show with default line style, or background = {None: cocopp.archives.bbob_mixint.get_all('')} for all algorithms benchmarked on the bbob-mixint test suite.

background_default_style =

Undocumented

balance_instances: bool =

give even weight to all instances by added columns with copies of underrepresented instances in DataSet.evals

dim_related_colors: tuple[str, ...] =

Undocumented

dim_related_markers: tuple[str, ...] =

Undocumented

extraction_folder_prefix: str =

Undocumented

figsize: list[float] =

Undocumented

figure_file_formats: list[str] =

Undocumented

filter_data_condition =

a callable which is used in rungenericmany to filter the list of DataSet. For example, to keep only the "effectively unimodal" functions of the 'bbob' testbed, set filter_data_condition = lambda ds: ds.funcId in [1, 2,] + list(range(5, 15))

force_assertions: bool =

Undocumented

foreground_algorithm_list: list =

a list of data files/folders as those specified in cocopp.main

in_a_hurry: int =

Undocumented

index_html_file_name: str =

Undocumented

inputCrE: float =

Undocumented

inputsettings: str =

Undocumented

instancesOfInterest =

Undocumented

instancesOfInterest2009: dict[int, int] =

Undocumented

instancesOfInterest2010: dict[int, int] =

Undocumented

instancesOfInterest2012: dict[int, int] =

Undocumented

instancesOfInterest2013: dict[int, int] =

Undocumented

instancesOfInterest2015: dict[int, int] =

Undocumented

instancesOfInterest2016: dict[int, int] =

Undocumented

instancesOfInterest2017: dict[int, int] =

Undocumented

instancesOfInterest2018: dict[int, int] =

Undocumented

instancesOfInterest2021: dict[int, int] =

Undocumented

instancesOfInterest2023: dict[int, int] =

Undocumented

instancesOfInterestBiobj2016: dict[int, int] =

Undocumented

instancesOfInterestBiobj2017: dict[int, int] =

Undocumented

interactive_mode: bool =

Undocumented

isConv: bool =

Undocumented

isExpensive: bool =

Undocumented

isFig: bool =

Undocumented

isLogLoss: bool =

Undocumented

isNoiseFree: bool =

Undocumented

isNoisy: bool =

Undocumented

isPickled: bool =

Undocumented

isRLDistr: bool =

Undocumented

isRldOnSingleFcts: bool =

Undocumented

isScatter: bool =

Undocumented

isTab: bool =

Undocumented

latex_commands_for_html: str =

Undocumented

len_of_names_in_pprldmany_legend =

set the length, for example when we want to remove characters that are not fully displayed, 9 == len('best 2009')

line_style_mapping: dict =

map the input argument position to a line style position, by default the identity. This could be useful to get the same style for several algorithm variants. Only effective with when bool(parameter_sweep) is True.

line_styles: list[dict] =

Undocumented

many_algorithm_file_name: str =

Undocumented

marker_size_multiplier: float =

multiply all marker sizes in below line_styles, used in compall.pprldmany.main

maxevals_fix_display =

Undocumented

minmax_algorithm_fontsize: list[int] =

Undocumented

minor_grid_alpha_in_pprldmany: float =

used in pprldmany for empirical runtime distributions, 0 means no minor grid

outputdir: str =

Undocumented

parameter_sweep: bool =

configure line styles for a parameter sweep, see also cocopp.config.config_line_styles

parameter_sweep_colormaps =

colormap(s) for parameter sweep set from the respective input long_option.

colormaps can, for example, be 'plasma.0.9,viridis' indicating the map names separated by commata and optionally the range for each map.

parameter_sweep_sort: bool =

assign color in sweep by leading float of algorithm name

ppconv_file_name: str =

Undocumented

ppfigcons1_file_name: str =

Undocumented

ppfigcons_file_name: str =

Undocumented

ppfigdim_file_name: str =

Undocumented

ppfigs_file_name: str =

Undocumented

pprldistr2_file_name: str =

Undocumented

pprldmany_file_name: str =

Undocumented

pprldmany_group_file_name: str =

Undocumented

ppscatter_file_name: str =

Undocumented

pptables_file_name: str =

Undocumented

rcaxes: dict[str, int] =

Undocumented

rcaxeslarger: dict =

Undocumented

rcfont: dict[str, int] =

Undocumented

rcfontlarger: dict[str, int] =

Undocumented

rclegend: dict[str, int] =

Undocumented

rclegendlarger: dict[str, int] =

Undocumented

rctick: dict[str, int] =

Undocumented

rcticklarger: dict[str, int] =

Undocumented

reference_algorithm_styles =

Undocumented

runlength_based_targets: bool =

Undocumented

scaling_figures_with_boxes: bool =

Undocumented

scaling_plots_with_axis_labels: bool =

Undocumented

sequential_colormaps: list[str] =

default color maps for a parameter sweep

simulated_runlength_bootstrap_sample_size =

bootstrap samples, 30 is a multiple of 10 and 15. Used for tables and plots. int(1e4) would be preferable for a final camera-ready paper version.

single_algorithm_file_name: str =

Undocumented

single_runlength_factors =

Undocumented

target_runlength: int =

Undocumented

target_runlengths_in_scaling_figs: list =

Undocumented

target_runlengths_in_single_rldistr: list =

Undocumented

target_runlengths_pprldmany =

Undocumented

target_runlengths_ppscatter =

Undocumented

test: bool =

Undocumented

use_recommendations: list[bool] =

use only recommendations data (.mdat files) for the respective algorithm where the last element is recycled for the remaining algorithms

verbose: bool =

Undocumented

warning_level: int =

Undocumented

weight_evaluations_constraints: list[int] =

weights used to sum function evaluations and constraints evaluations in the attribute DataSet._evals when data are loaded.

xlimit_expensive: float =

maximal run length multiplier in expensive setting, used in config for pprldmany and ppfigdim

xlimit_pprldmany: float =

maximal run length multiplier used in pprldmany.

This sets cocopp.compall.pprldmany.x_limit via calling cocopp.config.config()`.

Noisy setting should be rather 1e8? ?maybe better: 10 * genericsettings.evaluation_setting[1]?

_current_args =

arguments found in rungeneric.main, namely a list of folders returned by COCODataArchive.get_extended of cocopp.archives.all.

_default_line_styles =

Undocumented