eoDist: Total distance between observed and (closest) expected peaks.

Description Usage Arguments Details Author(s)

View source: R/functions_v9.r

Description

Calculates the total distance between observed peaks and (closest) expected peaks, with added penalty to force a specific copy number configuration to one of the peak. Not meant to be called directly.

Usage

1
2
eoDist(selectedPeaks, S, t, wd, xonly = F, weightbydiff = F, 
 weightbyarp = F, penaltymultiplier = 1000, penaltymultipliery = 0, ...)

Arguments

selectedPeaks

Passed from peakProximity

S

The S parameter

t

The t parameter (sums to 1).

wd

Weights given to each point in a cluster of points

xonly

Ignore the y-axis, only calculate distance based on the x-axis

weightbydiff

Ignore, not extensively tested.

weightbyarp

Ignore, not extensively tested.

penaltymultiplier

If a peak (a line of selectedPeak) is to be forced to represent a particular copy number configuration, then the fit is forced by adding a penalty multiplier to the distance between that peak and the expected location of the copy number configuration.

penaltymultipliery

If the penalty multiplier is to differ in the x- and y- directions, then use this as an additional multiplier in the y- direction (the penalty multiplier in the y-direction is then penaltymultiplier*penaltymultipliery).

...

Arguments passed to prepAR (ie, preserveMatPatDiff)

Details

To force a peak, a third column ($w) in selectedPeak is needed. That column contains a single non-zero integer value taking the format xy (for a 2-subclone model), where x is the copy number in the first sub-clone, y in the second subclone. In a 3-subclone system, the format is xyz, etc. All integers x,y,z,etc must take non-negative values no larger than 9.

Author(s)

Mathieu Lemire


mathieu-lemire/celluloid_0.11 documentation built on Aug. 6, 2018, 9:15 a.m.