plotPairs: Plot pairwise probability of co-membership

In maoyinan/BayesPC: R package for Bayesian Projection Clustering

Description

Given co-membership probability input based on postPairs, create a linkage plot colored according to probability range.

Usage

plotPairs(
  dat,
  id_var,
  g_var,
  ls_prob,
  ls_groups = NULL,
  ls_labels = NULL,
  cutPoints = c(0.5, 0.8),
  subTitles = paste(c("All", "Low", "Medium", "High"), "frequencies"),
  allColors = c("darkseagreen", "coral3", "steelblue", "sandybrown"),
  seed = 1,
  thin = 0.1
)

Arguments

dat	Longitudinal data input
id_var	Character of id variable
g_var	Column name in dat containing true group labels in the dataset
ls_prob	List of pairwise probability output of function postPairs
ls_groups	List of "similar" groups as elements of list respective to groups as list names. The co-membership probabilities indicated "similar" groups will be plotted in color allColors, and "dissimilar" groups will be plotted in black
ls_labels	List of group labels to display sequentially on the plot. Default is alphabetically sorted groups
cutPoints	Vector of two cutoff points between 0 and 1 to discretize probabilities into three intervals to indicate three levels of transparency of the curves. Default c(.5,.8)
subTitles	Vector of titles to be labeled for each subplot of random effect scenario
allColors	Colors of the curves under the random effect scenarios. Default c('darkseagreen', 'coral3','steelblue','sandybrown')
seed	Random seed for sampling curves when thin is less than 1
thin	Thinning parameter between 0 and 1. Recommended to set at low value to aid visualization. When thin=1 all curves are shown. Default is 0.1

Examples

ls_prob <- out_pc$ls_prob
ls_labels <- rep(list(1:4),4)
ls_labels[[4]] <- c(1,3,2,4)
ls_groups <- list(
'1'= c('2','3'),
'2'= c('1','4'),
'3'= c('1','4'),
'4'= c('3','2')
)
plotPairs(DATASET,'ID','Group',ls_prob,ls_groups,ls_labels,thin=.01)

maoyinan/BayesPC documentation built on Dec. 21, 2021, 1:48 p.m.