Endpointer function

Source: R/endpoint.R

The epR function is a tool for the identification of humane endpoints using single outcome variables from laboratory animal experiments. Originally, it was developed for using body weight but as values are normalized other continous variables are suited as well. The algorithm is highly functional in identifying strong deviations from a windowed normality. The hypothesis behind this is: larger deviations always point to severity.

epR(td = td, org = FALSE, wl = 6, SDwdth = 2, mad = FALSE,
  ltype = "b", dotcolor = "black", uprcol = "darkgreen",
  lwrcol = "magenta", cex = 1, cex.axis = 1, cex.lab = 1,
  xlim = NULL, ylim = NULL, pch = 19, blind = FALSE,
  ignupr = FALSE, xlab = "time", ylab = "Moving average (%)",
  main = NULL)

Arguments

td

testdata data.frame with n unique rows and p subsequent time points (e.g. days)
org

boolean (TRUE/FALSE) for using original values. If FALSE, data are normalized.
wl

SD window length (default is 6)
SDwdth

width of the standard deviation around the moving average (default is 2.5)
mad

boolean - use mean absolute deviation as quasi-clinical scoring constraint (default FALSE)
ltype

line type in shown plot
dotcolor

color of the shown dots (default "black")
uprcol

color of the upper threshold violation indicators
lwrcol

color of the upper threshold violation indicators
cex

point size
cex.axis

axis tick size
cex.lab

label size
xlim

range of x-axis (if set to NULL (default), plot will adapt automatically to given range - may not be nice!)
ylim

range of <-axis (if set to NULL (default), plot will adapt automatically to given range - may not be nice!)
pch

R-specific plot symbol for shown dots (default is 19)
blind

boolean (TRUE/FALSE) - if set to TRUE, no plot will be shown (default is FALSE)
ignupr

boolean (TRUE/FALSE) - ignore upper threshold violations (default is FALSE)
xlab

x-axis label
ylab

y-axis label
main

title

Value

data.frame with enpointeR results (n=number of data points, timepoint=index of marked endpoint, where=upper or lower boundary)

Examples

epR(as.numeric(gliodat[1,3:length(gliodat[1,])]), blind=TRUE )
#>    n timepoint value where
#> 1 29        12 99.25 lower
#> 2 29        15 94.14 lower