panel.meanspag: Lattice panel function for spagghetti plots with a mean line

In qPharmetra/qpToolkit: qPharmetra R Tools

Lattice panel function for spagghetti plots with a mean line

Description

Lattice panel function for spagghetti plots with a mean line

Usage

panel.meanspag(
  x,
  y,
  subscripts,
  groups,
  ...,
  myFun = mean,
  sumText = "Mean",
  individual.col = gray[5],
  type = "l",
  average.lwd = 3,
  average.col = red[8],
  add.legend = TRUE,
  legend.panel.row,
  legend.panel.column,
  legend.x,
  legend.y,
  show.points = FALSE
)

Arguments

x, y	The regular x, y input for panel functions
subscripts	default lattice pass-through
groups	default lattice pass-through
...	passed to panel.superpose()
myFun	the summarization function (mean)
sumText	text to describe the summary function ("mean")
individual.col	= color of individual lines (spagghetti elements)
type	like the lattice argument type for line type
average.lwd	line width of average line
average.col	color of average line
add.legend	logical (T) determining if a legend should be drawn
legend.panel.row	row number of the legend in the x by x grid of panels
legend.panel.column	column number of the legend in the x by x grid of panels
legend.x	x coordinate within the panel
legend.y	y coordinate within the panel
show.points	logical indicating to show observed data dots

Value

Lattice panel output (invisible)

Examples

## Not run: 
library(lattice)
pkpdData = example.pkpdData()
xyplot(value ~ time | dose *  type
       , data = pkpdData
       , subset = dose > 0 & value > 0
       , groups = id
       , aspect = 1
       , panel = panel.meanspag
)

out = get.xpose.tables(path = file.path(getOption("qpExampleDir")), run =  "example2")
trellis.strip.color()
xyplot(CONC ~ TIME
       , data = out
       , groups = ID
       , subset = EVID == 0
       , aspect = 1
       , scales = list(x = list(relation = "free"), y = list(log = 10))
       , panel = panel.meanspag
       , yscale.components = yscale.components.log10.3
)

# in case there are too many times to average by then the following might be #'#better interpretable
ptimes = seq(min(out$TIME),max(out$TIME), length = 14)
xyplot(CONC ~ Hmisc::Cbind(TIME, metrumrg::snap(TIME, rule = ptimes))
      , data = out
      , groups = ID
      , subset = EVID == 0
      , aspect = 1
      , scales = list(x = list(relation = "free"), y = list(log = 10))
      , panel = function(x,y,...)
      {
        panel.xyplot(x,y,..., col = gray[5], type = 'l')
        yy = tapply(y, attr(x, "other")[,1], mean)
        llines(as.numeric(names(yy)), yy, col = red[8], lwd = 2)
     }
      , yscale.components = yscale.components.log10.3
)

## End(Not run)

qPharmetra/qpToolkit documentation built on May 24, 2023, 8:52 a.m.