test.pairwise: Appnds formatted results of pairwise tests to data frame
In jakobbossek/tblutils: Helpers to pimp scientific tables

Description Usage Arguments Value

View source: R/test.pairwise.R

Split by some columns, further split by “split.col”, and do pairwise comparisons of the value.col column based on a statistical test. Results are stored in testresult.col.

test.pairwise(x, by, split.col, value.col, testresult.col,
  show.positive.only = FALSE, show.numbers.only = FALSE,
  colors = "black", show.intervals = FALSE, alpha = 0.05,
  test.fun = stats::wilcox.test, ...)

`x`	[`data.frame`] Data frame.
`by`	[`character`] Column names used for splitting the data frame. This will usually be columns that describe the problems / problem parameters.
`split.col`	[`character(1)`] Single column name. This will usually be the algorithm column.
`value.col`	[`character(1)`] Single column name with numeric values to undergo statistical tests.
`testresult.col`	[`character(1)`] Column name of new column where test results should be stored.
`show.positive.only`	[`logical(1)`] Show entries for rejected zero hypothesis only? Default is `FALSE`.
`show.numbers.only`	[`logical(1)`] Show algorithm number only, i.e., without plus/minus indicator for test result? The value `TRUE` is only meaningful if `show.positive.only = TRUE`. Default is `FALSE`.
`colors`	[`character`] Vector of colors used to highlight test results. If one color is passed it is simply repeated k times where k is the number of algorithms. Otherwise the k-th color is used for latex output if the zero hypothesis is rejected for this algorithm.
`show.intervals`	[`logical(1)`] Show results in intervals when possible and not comma-seperated? This might lead to narrower columns sometimes. Default is `FALSE`.
`alpha`	[`numeric(1)`] Significance level. Default is 0.05, i.e. five percent.
`test.fun`	[`function(x, y, ...)`] Statistical test function. Default is `wilcox.test`.
`...`	[any] Further parameters passed down to `test.fun`. E.g., `alternative = "less"`.