fitNorm2 {prada}R Documentation

Fit bivariate normal distribution.

Description

Fits a bivariate normal distribution into a data set of paired values and selects data points according to their standard deviation from the fitted distribution.

Usage

fitNorm2(x, y=NA, scalefac=1, method="covMcd", noise, gateName = "fitNorm")

Arguments

x Numeric vector containing x-value or n by 2 matrix containing x and y values or object of class cytoFrame.
y Numeric vector containing y-value (optional). The length of x must be the same as that of y.
scalefac Numeric vector giving factor of standard deviations used for data selection (all points within scalefac standard deviations are selected).
method One of covMcd or cov.rob defining method used for computation of covariance matrix.
noise Numeric or logical index vector defining value pairs in x that are not used for fitting of distributions. Can be used to deal with noisy data.
gateName Character giving the name of the gate object.

Details

Computes the densities of a bivariate normal distribution from the covariance matrix of the paired data. Covariance matrices are acquired either by function covMcd (considerably faster) or by function cov.rob.

Value

A list containing items mu (midpoint of distribution), S (covariance matrix), p (density values for each data pair), sel (selection of data points), scalefac (factor of standard deviations used for data selection), data (x and y values of data points) and gate, an object of class gate containing the selection.

Author(s)

Florian Hahne

See Also

cov.rob, covMcd, plotNorm2

Examples

 sampdat <- readFCS(system.file("extdata",
   "fas-Bcl2-plate323-04-04.A01", package="prada"))
 nfit    <- fitNorm2(exprs(sampdat[,1:2]), scalefac=2)
 plotNorm2(nfit, selection=TRUE, ellipse=TRUE) 

[Package prada version 1.12.0 Index]