lattice-methods: Methods implementing Lattice displays for flow data
In flowViz: Visualization for flow cytometry

Description Usage Arguments Details Methods Examples

Various methods implementing multipanel visualizations for flow data using infrastructure provided in the lattice package. The original generics for these methods are defined in lattice, and these S4 methods (mostly) dispatch on a formula and the data argument which must be of class flowSet or flowFrame. The formula has to be fairly basic: conditioning can be done using phenodata variables and channel names (the colnames slot) can be used as panel variables. See examples below for sample usage.

## S4 method for signature 'formula,flowSet'
levelplot(x, data, xlab, ylab,
  as.table = TRUE, contour = TRUE, labels = FALSE, n = 50, ...)

## S4 method for signature 'formula,flowSet'
qqmath(x, data, xlab, ylab,
  f.value = function(n) ppoints(ceiling(sqrt(n))),
  distribution = qnorm, ...)

## S4 method for signature 'flowFrame,missing'
parallel(x, data, reorder.by = function(x)
  var(x, na.rm = TRUE), time = "Time", exclude.time = TRUE, ...)

## S4 method for signature 'formula,flowSet'
parallel(x, data, time = "Time",
  exclude.time = TRUE, filter = NULL, xlab = NULL, ylab = NULL,
  ...)

`x`	a formula describing the structure of the plot and the variables to be used in the display.
`data`	a `flowSet` object that serves as a source of data.
`xlab, ylab`	Labels for data axes, with suitable defaults taken from the formula
`as.table, contour, labels`	These arguments are passed unchanged to the corresponding methods in lattice, and are listed here only because they provide different defaults. See documentation for the original methods for details.
`n`	the number of bins on each axis to be used when evaluating the density
`...`	more arguments, usually passed on to the underlying lattice methods.
`f.value, distribution`	number of points used in Q-Q plot, and the reference distribution used. See `qqmath` for details.
`reorder.by`	a function, which is applied to each column. The columns are ordered by the results. Reordering can be suppressed by setting this to `NULL`.
`time`	A character string giving the name of the column recording time.
`exclude.time`	logical, specifying whether to exclude the time variable from a scatter plot matrix or parallel coordinates plot. It is rarely meaningful not to do so.
`filter`	flowCore filter

Not all standard lattice arguments will have the intended effect, but many should. For a fuller description of possible arguments and their effects, consult documentation on lattice (Trellis docs would also work for the fundamentals).

qqmath: signature(x = "formula", data = "flowSet"): creates theoretical quantile plots of a given channel, with one or more samples per panel
levelplot: signature(x = "formula", data = "flowSet"): similar to the xyplot method, but plots estimated density (using kde2d) with a common z-scale and an optional color key.
parallel: signature(x = "flowFrame", data = "missing"): draws a parallel coordinates plot of all channels (excluding time, by default) of a flowFrame object. This is rarely useful without transparency, but that is currently only possible with the pdf device (and perhaps the aqua device as well).

library(flowCore)
data(GvHD)

qqmath( ~ `FSC-H` | factor(Patient), GvHD,
       grid = TRUE, type = "l",
       f.value = ppoints(100))


## contourplot of bivariate density:

require(colorspace)
YlOrBr <- c("#FFFFD4", "#FED98E", "#FE9929", "#D95F0E", "#993404")
colori <- colorRampPalette(YlOrBr)
levelplot(asinh(`SSC-H`) ~ asinh(`FSC-H`) | Visit + Patient, GvHD, n = 20,
          col.regions = colori(50), main = "Contour Plot")



## parallel coordinate plots

parallel(GvHD[["s6a01"]])

## Not run: 

## try with PDF device
parallel(GvHD[["s7a01"]], alpha = 0.01)


## End(Not run)