lad.fit.BR

lad.fit.EM

Fits a linear model using LAD methods, returning the bare minimum computations.

regression

lad.fit-methods

L1 estimation for linear regression using Barrodale and Roberts' method
<doi:10.1145/355616.361024> and the EM algorithm <doi:10.1023/A:1020759012226>.
Estimation of mean and covariance matrix using the multivariate Laplace distribution,
density, distribution function, quantile function and random number generation
for univariate and multivariate Laplace distribution <doi:10.1080/03610929808832115>.
Implementation of Naik and Plungpongpun <doi:10.1007/0-8176-4487-3_7> for the
Generalized spatial median estimator is included.

Felipe Osorio

L1pack

Routines for L1 Estimation

Tymoteusz Wolodzko

lad.fit-methods function

<dl> <dt>x, y</dt>
<dd>numeric vectors or matrices for the predictors and the response in
 a linear model. Typically, but not necessarily, <code>x</code> will be constructed by
 one of the fitting functions.</dd> <dt>tol</dt>
<dd>the relative tolerance for the iterative algorithm. Default is <code>tol = 1e-7</code>.</dd> <dt>maxiter</dt>
<dd>The maximum number of iterations for the <code>EM</code> method. Default to 200.</dd></dl>

Arguments

Fit a least absolute deviation (LAD) regression model — lad.fit-methods

<dl>

 <dt>x, y</dt>
<dd>numeric vectors or matrices for the predictors and the response in
 a linear model. Typically, but not necessarily, <code>x</code> will be constructed by
 one of the fitting functions.</dd>

 <dt>tol</dt>
<dd>the relative tolerance for the iterative algorithm. Default is <code>tol = 1e-7</code>.</dd>

 <dt>maxiter</dt>
<dd>The maximum number of iterations for the <code>EM</code> method. Default to 200.</dd>

</dl>

lad.fit-methods: Fit a least absolute deviation (LAD) regression model

Description

Usage

Value

Arguments

See Also

Examples