classify: Classifies normalised AFLP data
In AFLP: Analysis of AFLP data

Description Usage Arguments Details Value Author(s) See Also Examples

The normalised fluorescence score of AFLP data are converted into one or more classes. The classification is based on the first and second derivative of the density of the normalised fluorescence. Hence the data must be normalised prior to classification.

  classify(data, output = c("screen", "none", "tex"),
    maxBorder = 1, tresholdPeakRatio = 0.03,
    tresholdMonomorph = 0, tresholdMaxValley = 0.95,
    device = "pdf", path = NULL, nrows = 4, ncols = 4,
    keep.border = FALSE)

`data`	A normalised AFLP object. Unnormalised object will be normalised prior to classification.
`output`	Indicates how the inspections plots should be given. "none" suppresses them, "screen" prints them on the screen, "tex" save them to a file and generates LaTeX to include these files into a document.
`maxBorder`	The maximum number of borders per marker.
`tresholdPeakRatio`	The minimum height of a potential peak in the density relative to the maximum density. Peak with a ratio lower that this threshold are not considered as peaks. Defaults to 0.03.
`tresholdMonomorph`	Calculate the ratio of 'absent' and 'present' bins in a marker. Consider the marker as monomorph if either of them is smaller than the threshold. Defaults to 0.
`tresholdMaxValley`	A treshold for valleys between peaks in the relative density. P,my vallyes below this treshold are considered for breaks. Defaults to 0.95.
`path`	the path where the figures are saved. Only used if `output = "tex"`. Defaults to NULL, which is the working directory.
`device`	the device to which the figures are saved. See `ggsave` for the available devices. Only used if `output = "tex"`. Defaults to "pdf".
`nrows`	Gives the prefer number of rows with plot per figure. Defaults to 4.
`ncols`	Gives the prefer number of columns with plot per figure. Defaults to 4.
`keep.border`	Logical. If FALSE then the borders are recalculated and overwritten. If TRUE the borders in the object are used for the classification. The latter is intended to update the classification after manually altering the borders. Default to FALSE.

Monomorph markers will have only one class and the corresponding border with have Inf as cut-level. The lowest class in polymorphic markers is considered an 'absent' locus, all other classes indicate a 'present' locus.

We strongly suggest that the user thoroughly inspects the resulting plots as the automated classification might yield some unwanted artifacts. Correct them with the link{border<-} function.

Returns the data object with a updated borders slot and updated Score in the Fluorescence slot.

Thierry Onkelinx Thierry.Onkelinx@inbo.be, Paul Quataert

border<-, border, normalise, ggsave