rrcov-utils: Different utility functions to be used in rrcov and packages...

In rrcov: Scalable Robust Estimators with High Breakdown Point

Different utility functions to be used in rrcov and packages depending on rrcov

Description

Different utility functions to be used in rrcov and packages depending on rrcov:

• myscreeplot Plots screeplots for two covariance matrices

• mtxconfusion Computes a confusion matrix between actual and predicted class membership

• MATLAB-like functions

  • sqrtm Returns the square root of a symetric positive definite matrix

  • ones Returns a nxp matrix filled with ones

  • zeros Returns a nxp matrix filled with zeros

  • vecnorm Returns the norm of a vector

  • rankMM Returns the rank of the matrix

  • repmat Replicates a matrix

Usage

    myscreeplot(rcov, ccov)
    mtxconfusion(actual, predicted, prior = NULL, printit=FALSE)
    sqrtm(A)

Arguments

A	A matrix.
tol	Optional numerical tolerance (compared to singular values).
sv	A vector of non-increasing singular values of a matrix A, if already computed.
rcov, ccov	Covariance matrices.
n	Number of rows.
p	Number of columns.
actual	Actual class membership.
predicted	Predicted class membership.
prior	Prior probabilities.
printit	Wheather to print the results.

Author(s)

Valentin Todorov [e-mail]

rrcov documentation built on July 9, 2023, 6:03 p.m.