multiRunPlot_2: Plot the results from multipe COBRA runs.

In SACOBRA: Self-Adjusting COBRA

Description

Plot for each run one black curve 'error vs. iterations' and aggregate the mean curve (red) and the median curve (green) of all runs. DIFFERENCE to multiRunPlot: 'error' is the distance of the ever-best feasible point in input space to the true solution solu.

Usage

multiRunPlot_2(
  dfAll,
  solu,
  fName = "multiRun",
  main = "",
  xlim = NULL,
  ylim = c(1e-05, 10000),
  ylog = TRUE,
  xlog = FALSE,
  target = 0.05,
  plotPDF = FALSE,
  subPDF = NULL,
  legendWhere = "topright",
  absErr = FALSE
)

Arguments

dfAll	the data frame of all runs, obtained with multiCOBRA or loaded from .Rdata file
solu	the true solution in input space of the problem (only for diagnostics).
fName	["multiRun"] the name of the .Rdata file, printed as subtitle
main	[""] the name of the problem (e.g. "G01 problem"), printed as title
xlim	the x limits
ylim	the y limits
ylog	[TRUE] logarithmic y-axis
xlog	[FALSE] logarithmic x-axis
target	[0.05] a single run meets the target, if the final error is smaller than target
plotPDF	[FALSE] if TRUE, plot to <fName>.pdf
subPDF	[NULL] optional subdirectory where .pdf should go
legendWhere	["topright"]
absErr	[FALSE] if TRUE, plot abs(error) instead of error.

Details

Print some diagnostic information: final median & mean error, percentage of runs which meet the target (only if optim is available)).

Value

z3, a vector containing for each run the ever-best feasible objective value

Author(s)

Wolfgang Konen, Samineh Bagheri, Cologne University of Applied Sciences

See Also

multiRunPlot, multiCOBRA, cobraPhaseII

SACOBRA documentation built on March 26, 2020, 7:15 p.m.