plot.bim.diag {bim}R Documentation

Marginal and model-conditional summaries of Bayesian interval mapping diagnostics

Description

A density histogram is drawn for model-averaged summary diagnostics such as LOD, variance, or heritability.

Usage

plot.bim.diag(x, nqtl=1, pattern=NULL, exact=FALSE,
  items=<<see below>>, mains=items,
  mfrow=<<see below>>, ... )

Arguments

x object of class bim
nqtl subset on number of QTL
pattern subset on chromosome pattern of QTL
exact subset on exact pattern or number of QTL if true
items diagnostics to be summarized; must be column of data
mains titles for items
mfrow plot arrangement parameter for par() (default is rows = number of items by cols = 2)
... graphical parameters can be given as arguments to plot

Details

Model-averaged density is smooth kernel estimate similar to ordinary histogram. A boxplot (without outliers) is overlaid for comparison with conditional boxplots. Conditional boxplots by number of QTL may show indication of model bias for small number of QTL. This and bim.nqtl can help suggest the minimal model. Diagnostic items that make sense to plot are "LOD", "envvar" (environmental variance), "herit" (heritability), "mean" (grand mean), "addvar" (variance of add), "domvar" (variance of add). Marginal and conditional medians are printed.

Author(s)

Brian S. Yandell, yandell@stat.wisc.edu

References

http://www.stat.wisc.edu/~yandell/qtl/software/Bmapqtl

See Also

plot.bim, density, boxplot

Examples

data( verngeo.bim )

plot.bim.diag( verngeo.bim, 2, items = c("LOD","herit") )

[Package bim version 1.01-1 Index]