happyplot {happy} | R Documentation |
happyplot() will plot along the genome the log P-value that a QTL is not found in a series of marker intervals. It accepts as input the results of hfit(), mfit() and mergefit(). mergeplot() is a convenience function for calling happlyplot() after a call to mergefit(), with several parameters set.
happyplot( fit, mode='logP', labels=NULL, xlab='cM', ylab=NULL, main=NULL, t='s', pch=20, ... ) mergeplot( fit, mergedata, mode='logP', xlab='bp', ylab=NULL, main=NULL, t='p', pch=20, ... )
fit |
an object returned by a previous call to hfit(), mfit(), or mergefit() |
mode |
the mode of the plot - either 'logP', when the negative base-10 logarithm of the ANOVA P-value of plotted, or 'SS', when the fitting sums-of-squares is plotted |
labels |
optional matrix detailing marker labels to be drawn on the plot. The labels are written vertically above the plot, with vertical lines extending down into the plot area. labels is a matrix with two named columns 'marker', containing the marker names, and 'POSITION', containing the x-axis positions of the markers. |
mergedata |
(mergeplot() only). an object returned by a previous call to mergeprepare(). This is used to construct labels for plotting |
xlab |
the x-axis label |
ylab |
the y-axis label |
main |
the titke of the plot |
t |
the type of plot - either 'p', 'l', 's' or 'S', with the same meanings as in plot() |
pch |
the plotting character code, with the same meaning as in plot() |
... |
arguments passed down to R plotting functions |
A plot to the current graphics device is produced. For happyplot, if fit$permdata is not NULL (i.e. hfit() was run using the permutation test option) then the plot contains two step functions, labelled global.logp and point.logp The global.logp plot shows the empirical log-pvalue relative the whole region (ie adjusted for the number of markers) while point.logp shows the empirical log-pvalue for each interval. If fit$permdata is NULL then the plots give the ANOVA logP values. If the model used in hfit() is 'additive' then the logP for the additive mocdel vs the null model is plotted; if the model is 'full' then the curves for the full, additive and partial F-test logP values are plotted.
Richard Mott
hfit(), mfit(), mergefit()
## Not run: h <- happy( 'HS.data', 'HS.alleles' ) ## Not run: fit <- hfit( h, h$markers, model='full' ) ## Not run: happyplot( fit )