aitoa.plot.progress.stat.on.multiple.instances: Plot the Progress of a Statistic for a Set of Algorithms on a...

In thomasWeise/aitoaEvaluate: An R Package for Evaluating the aitoa Results

Description

Plot how a statistic for each element of set of algorithms progress over a set of problem instances. For each instance, one diagram is plotted. The diagrams are arranged one by one from in a vertical row.

Usage

aitoa.plot.progress.stat.on.multiple.instances(
  results.dir,
  algorithms,
  instances,
  time.column = c("t", "fes"),
  max.time = NA_integer_,
  algorithm.colors = aitoa.distinct.colors(length(algorithms)),
  quantiles = .stat.plot.default.quantiles,
  center.stat = median,
  center.lty = .default.lty,
  center.lwd = .thick.lwd,
  quantile.transparency = 0.8,
  make.stairs.quantiles = FALSE,
  make.stairs.center = FALSE,
  instances.limit = NA_integer_,
  instances.limit.name = NULL,
  instance.limit.color = .instance.limit.color,
  instance.limit.lty = .instance.limit.lty,
  instance.limit.lwd = .instance.limit.lwd,
  legend.cex = .legend.cex,
  legend.bg = .legend.bg,
  time.axis.text = if (time.column[[1L]] == "t") .time.ms.text else .time.fes.text,
  quality.axis.text = .quality.text,
  make.time.unique = (time.column[[1L]] == "t"),
  use.f.range.from.raw.data = FALSE,
  mgp = .default.mgp,
  tck = .default.tck,
  cex = .default.cex,
  mar = .default.mar.without.labels,
  mar.single = .default.mar.without.labels,
  ...
)

Arguments

results.dir	the directory where the results can be loaded from
algorithms	the list of algorithm IDs. The names of this list, if set, will be used in the legends.
instances	the list of instance IDs. The names of this list, if set, will be used in the legend
time.column	the time dimension, either t or fes
max.time	an optional limit for the time, can be a vector
algorithm.colors	a character vector of the same length as algorithms providing the colors to be used for the algorithms
quantiles	the quantiles to be plotted in a shaded fashion: all values must be >= 0 and < 0.5 and are mirrored. I.e., if you specify 0.2, then the region between the 0.2 and 0.8 quantile will be plotted semi-transparent.
center.stat	the central statistic to plot, usually the median or mean
center.lty	the line type to be used for the median lines
center.lwd	the line width to be used for median lines
quantile.transparency	the transparency to be applied to each quantile polygon
make.stairs.quantiles	should the quantiles be shown as stairs (TRUE) or linear interpolation (FALSE)
make.stairs.center	should the center statistic be shown as stairs (TRUE) or linear interpolation (FALSE)
instances.limit	an optional vector of lower bounds or best-known solutions for the instances
instances.limit.name	an optional name or name vector for the instances limits
instance.limit.color	the color for the instance limit line
instance.limit.lty	the line type for the instance limit line
instance.limit.lwd	the line width for the instance limit line
legend.cex	the character scaling for the legend
legend.bg	the background color for the legend
time.axis.text	the text to be used for labeling the time axis, NA for omit label
quality.axis.text	the text to be used for labeling the quality axis, NA for omit label
make.time.unique	should we make the time indices unique (except maybe for the first and last point)? This makes sense when we want to plot diagrams over a time axis, as we then have removed redundant points right away. If make.time.unique==FALSE, then there may be multiple improvements at the same time index due to the resolution of the computer clock (while each improvement will definitely have a unique FE).
use.f.range.from.raw.data	should we use the real data to compute the range of the y-axis, or should the range depend on the computed statistics only (default)?
mgp	the mgp parameter to be passed to plot
tck	the tck parameter to be passed to plot
cex	the default character scaling
mar	the default margins, can be a list of vectors, one for each plot
mar.single	the margin parameter(s) to be passed to the single diagrams
...	parameters to be passed to par

thomasWeise/aitoaEvaluate documentation built on Dec. 6, 2020, 1:22 p.m.