plotFitted function

Plot observed segregation ratios and fitted and theoretical models

Plots histogram of observed segregation ratios on logit scale along with scaled density of fitted components corresponding to dosage classes. Plots of expected theoretical distributions can be plotted with or without segregation ratio data.

## S3 method for class 'runJagsWrapper'
plot(x, theoretical=FALSE, ...)

plotFitted(seg.ratios, summary.mixture, add.random.effect=TRUE,
  theoretical=FALSE, model=NULL, theory.col="red",
  xaxis=c("logit","raw"), ylim=NULL, NCLASS=NULL, n.seq=100,
  xlab="logit(Segregation Ratio)", ylab="Density", density.plot=FALSE,
  fitted.lwd=2, fitted.col="blue", bar.col="lightgreen", cex=1,
  warnings = FALSE, main=NULL, ...)

plotTheoretical(ploidy.level=8, seg.ratios=NULL, n.components=NULL,
  expected.segratio=NULL, proportions=c(0.65,0.2,0.1,0.03,0.01,0.01, 0, 0),
  n.individuals=200, xaxis=c("raw","logit"),
  type.parents=c("heterogeneous","homozygous"), xlim=c(0,1),
  NCLASS=NULL, xlab="Segregation Ratio", ylab="Density",
  density.plot=FALSE, fitted.lwd=2, fitted.col="blue", cex=1,
  warnings = TRUE, main=NULL, ...)

Arguments

• x: object of class runJagsWrapper produced by using runSegratioMM to set up and fit mixture model
• seg.ratios: segregation ratios as class segRatio
• summary.mixture: mcmc summary data produce by summary.segratioMCMC
• add.random.effect: add random variance component to fitted distribution plot if model includes a random effect (default: TRUE)
• theoretical: whether to plot the expected theoretical distribution under the fitted model (default: FALSE)
• model: object of class modelSegratioMM specifying model if plotting expected theoretical distribution
• theory.col: colour for expected theoretical distribution (default: "red")
• ploidy.level: the number of homologous chromosomes
• n.components: number of components for mixture model
• expected.segratio: may be specified or automatically calculated from ploidy level etc
• xaxis: whether to plot on "logit" or "raw" scale. Defaults to "logit" if plotting segregation ratios or "raw" for theoretical distributions
• proportions: for no. of markers in each component of theoretical distribution plot
• n.individuals: for theoretical distribution plot - taken from segregation ratios if supplied
• type.parents: "heterogeneous" if parental markers are 0,1 or "homogeneous" if parental markers are both 1
• ylim: c(lower,upper) yaxis limits for histogram of segregation ratios
• xlim: c(lower,upper) xaxis limits for segregation ratios
• NCLASS: number of classes for histogram (Default: 100)
• n.seq: number of points to use for plotting fitted mixture
• xlab: x-axis label
• ylab: y-axis label
• density.plot: whether to plot a smoothed density as well as segregation data and fitted and/or theoretical distributions (default: FALSE)
• main: title for plot
• fitted.lwd: width for fitted line
• fitted.col: colour for fitted line
• bar.col: colour for histogram
• cex: character expansion for text (see par)
• warnings: print warnings like number of components etc (Default: FALSE)
• ...: extra options for plot

Details

plotFitted plot histogram of observed segregation ratios on logit scale along with scaled density of fitted components corresponding to dosage classes using trellis

plotTheoretical plot expected distribution of autopolyploid dominant markers on probability (0,1) scale. Segregation ratios may also be plotted

plot.runJagsWrapper plots the fitted values of object of class runJagsWrapper which has been produced by using runSegratioMM to set up and fit mixture model

Note that since trellis graphics are employed, plots may need to be printed in order to see them

Returns

None.

Author(s)

Peter Baker p.baker1@uq.edu.au

See Also

summary.mcmc mcmc

segratioMCMC readJags

diagnosticsJagsMix runSegratioMM

Examples

## simulate small autooctaploid data set
plotTheoretical(8, proportion=c(0.7,0.2,0.1),n.individuals=50)
a1 <- sim.autoMarkers(8,c(0.7,0.2,0.1),n.markers=100,n.individuals=50)
##print(a1)
sr <-  segregationRatios(a1$markers)
x <- setModel(3,8)

## fit simple model in one hit and summarise
## Not run:

x.run <- runSegratioMM(sr, x, burn.in=200, sample=500)
print(x.run)

## plot fitted model using 'plotFitted'
plotFitted(sr, x.run$summary)
a.plot <- plotFitted(sr, x.run$summary, density.plot=TRUE)
print(a.plot)
## or the easier way
plot(x.run, theoretical=TRUE)
## End(Not run)