timeLinePlot: Plot channel values against time

Description

Plots values of one parameter for each flowFrame in a flowSet against time.

Usage

timelineplot(x, channel, type = c("stacked", "scaled", "native", "frequency"),
  col, ylab = names(x), binSize, varCut = 1, ...)
## S3 method for class 'flowSet,character':
timeLinePlot(x, channel, type = c("stacked",
  "scaled", "native", "frequency"), col = NULL, ylab = sampleNames(x),
  binSize, varCut = 1, ...)
## S3 method for class 'flowFrame,character':
timeLinePlot(x, channel, ...)
## S3 method for class 'ANY,missing':
timeLinePlot(x, channel, ...)

Arguments

An object of class flowFrame or flowSet containing the data to be plotted.

channel

The parameter for which the data is to be plotted

type

One in 'stacked', 'scaled' or 'native'. 'stacked' will plot the measurements for the frames on top of each other. 'scaled' will align the median values around zero and 'native' will plot the values in the original dimensions of the measurement range.

col

Optional color parameter.

ylab

The axis annotation to add on the y-axis for stacked plots.

binSize

The number of events per bin. If not set, a reasonable default is computed.

varCut

The cutoff in the adjusted variance to which the quality score is computed. Basically, all values that are outside of the confidence interval defined by [my - signma * varCut, my + sigma * varCut] will contribute to a positive quality score value.

...

Further arguments that are passed on to the base plotting functions.

Value

A numeric vector of quality scores.

Details

Plotting flow cytometry data against the time domain can help to identify problems with the fluidics or drifts in the instrument setting during measurement runs.

This function creates plots for all flowFrames in a flowSet for a given parameter against time. A barplot legend indicates the deviation from the median for each sample. There is also a flowFrame method, which will create a plot for a single flowFrame only.

In addition, the function computes a quality score for each frame, which essentially is the sum of the positive distances of each bin mean from a frame-specific confidence interval, divided by the number of bins. Values larger than zero indicate a problem.

Examples

Run this code

data(GvHD)
opar <- par(ask=TRUE)

res <- timeLinePlot(GvHD[[1]], "SSC-H")
res

res <- timeLinePlot(GvHD, "SSC-H")

res <- timeLinePlot(GvHD, "SSC-H", type="scaled", varCut=4)

res <- timeLinePlot(GvHD[1:4], "SSC-H", type="native", binSize=50)

par(opar)