oneVsAllPlot: Plotting all variables against a single variable

View source: R/oneVsAllPlot.R

oneVsAllPlotR Documentation

Plotting all variables against a single variable

Description

This function is useful both when setting appropriate gates and when the adjustments of the compensation are done

Usage

oneVsAllPlot(
  flowObj,
  yCol = "all",
  nRows = 10000,
  plotName = "default",
  dirName = "All_vs_all_plots",
  zeroTrim = TRUE,
  hexBins = 30,
  saveResult = TRUE
)

Arguments

flowObj

This is the full dataset, either a flowFrame or a flowSet, that should be plotted. If it has more rows than "nRows", a subsample (with equal contributions from each sample if a flowSet) will be plotted.

yCol

Here, the variable to be plotted against all the others is selected. It can be either a number, the column name of interest or "all".

nRows

The number of rows that will be used to construct the plot. The fewer, the faster, but the resolution also decreases, naturally. Default is 100000.

plotName

If a name different from yCol should be used, it can be added here.

dirName

Add a custom directory name.

zeroTrim

In the case of CyTOF data, the events at zero can often be so dominant, that all other density variation is dwarfed, and thus invisible. With this command, the events that are zero in both y and x are trimmed for each x separately.

hexBins

How many bins should the hexagonal plots be divided in?

saveResult

Should the result be saved as a file?

Value

A plot with one 2D-graph for each variable that the y-variable should be plotted against.

See Also

correctUnmix

Examples

#' # Load uncompensated data
data(fullPanel)

# Load the spectral unmixing matrix generated with controls from the same
# experiment. These can be generated using the specMatCalc function.
data(specMat)

# Now unmix
fullPanelUnmix <- specUnmix(fullPanel, specMat)

# Transform all fluorescent channels
fullPanelTrans <- arcTrans(fullPanelUnmix,
    transNames = colnames(fullPanelUnmix)[6:18])

# And now run the function. If no specific marker is selected, as in this
# case, then all markers will be plotted in a new sub-directory.
# Further, if you leave the saveResult to TRUE, a pdf will be created.
oneVsAllPlot(fullPanelTrans, yCol = "BV650_CD56", saveResult = FALSE)

# This shows that there is a compensation artifact between AF647_IgM and
# BV650_CD56, which is an expected combination to cause problems, due to the
# similar emission characteristics. It is therefore recommended to go on to
# correctUnmix function.


jtheorell/flowSpecs documentation built on April 24, 2023, 2:59 a.m.