drawAlgorithmGraph: Hypotheses represented as a graph

In scmamp: Statistical Comparison of Multiple Algorithms in Multiple Problems

Description

This function can be used to plot a graph where algorithms are nodes and algorithms that cannot be regarded as different are joined by an edge.

Usage

drawAlgorithmGraph(pvalue.matrix, mean.value, ..., alpha = 0.05,
  font.size = 15, highlight = "min", highlight.color = "chartreuse3",
  node.color = "gray30", font.color = "white", digits = 2,
  node.width = 5, node.height = 2)

Arguments

pvalue.matrix	Matrix with the p-values
mean.value	Vector of values to be written together with the name of the algorithm
...	Additional parameters to the Rgraphviz function. This is mainly to change the layout of the graph
alpha	Significance level to determine which hypotheses are rejected.
font.size	Size of the font for the node labels.
highlight	A character indicating which node has to be highlighted. It can be the one with the maximum value ('max'), the minimum value ('min') or none ('none').
highlight.color	Any R valid color for the highlighted node.
node.color	Any R valid color for the non-highlighted nodes.
font.color	Any R valid color for the node labels.
digits	Number of digits to display the value associated to each node
node.width	Numeric value for the width of the node
node.height	Numeric value for the height of the node

See Also

plotPvalues, plotRanking, plotCD

Examples

data(data_blum_2015)
data <- filterData(data.blum.2015, condition="Size == 1000", remove.cols=1:2)
res <- postHocTest(data, test = "friedman", use.rank=TRUE, correct="bergmann")
## This function requieres the package Rgraphviz
# drawAlgorithmGraph(res$corrected.pval, res$summary)

scmamp documentation built on May 1, 2019, 10:10 p.m.