sSeq (version 1.10.0)

plotDispersion: Drawing Dispersion-Mean plot.

Description

This function is used to draw a scatter plot of dispersion versus mean of count table. It helps to visually inspect the dependence between the dispersion estimates and the mean estimates.

Usage

plotDispersion(DispSH, extraOutput=NULL, plotMethod="logDisp", ylim1=NULL, legPos="topleft", myCol=brewer.pal(9, "Set1"),  tt=NULL)

Arguments

DispSH

A data frame includes `SH', `raw', and `mus'. They are the shrinkage estimates of dispersion, the method of moment estimates of dispersion, and the estimates of mean. This data frame is obtained using the function `nbTestSH' and specifying `SHonly=TRUE'.

extraOutput

A data.frame including dispersion estimates and expectation estimates using another method. When users want to compare the dispersion estimates using two different method, this argument can be used to include the result from the second method. The default value is NULL. This means that no extra method is compared.

plotMethod

If plotMethod=``logDisp" which is the default, then both dispersion and mean estimates are shown in the log scale. If plotMethod=``Disp", then only mean estimates are shown in the log scale.

ylim1

A vector of two values that specifies the minimum and maximum values of the vertical y axis in the plot. It is used to limit the presenting range of y axis in the plot. If ylime1=NULL then the range of the shrinkage estimates of dispersion is used.

legPos

A character indicating the position of legend in the plot. The value of this argument can be "topleft", "topright", "bottomleft" and "bottomright".

myCol

A vector of colors corresponding to the dispersions estimated using different methods.

tt

A character representing the title of the plot, which is shown on the top in the plot.

Examples

data("countsTable")
conds <- c("A", "B")
dispSH <- nbTestSH( countsTable, conds, "A", "B", SHonly=TRUE)

library('RColorBrewer')
plotDispersion(dispSH, legPos="topleft")