normalizeData {pcFactorStan} | R Documentation |
Normalize data according to a canonical order
Description
Pairwise comparison data are not commutative.
Alice beating Bob in chess is equivalent to Bob losing to
Alice. normalizeData
assigns an arbitrary order to all
vertices and reorders vertices column-wise to match,
flipping signs as needed.
Usage
normalizeData(df, ..., .palist = NULL, .sortRows = TRUE)
Arguments
df |
a data frame with pairs of vertices given in columns |
... |
Not used. Forces remaining arguments to be specified by name. |
.palist |
a character vector giving an order to use instead of the default |
.sortRows |
logical. Using the same order, sort rows in addition to vertex pairs. |
Examples
df <- data.frame(pa1=NA, pa2=NA, i1=c(1, -1))
df[1,paste0('pa',1:2)] <- c('a','b')
df[2,paste0('pa',1:2)] <- c('b','a')
normalizeData(df)
[Package pcFactorStan version 1.5.4 Index]