ogls: Omnivariant Generalised Least-Squares Regression

In IsoplotR: Statistical Toolbox for Radiometric Geochronology

Omnivariant Generalised Least-Squares Regression

Description

Linear regression with inter-sample error correlations.

Usage

ogls(x, ...)

## Default S3 method:
ogls(x, random.effects = FALSE, ...)

## S3 method for class 'other'
ogls(x, random.effects = FALSE, ...)

Arguments

x	either a n x (n+1) matrix obtained by prepending a vector of alternating X,Y-values to its covariance matrix OR an object of class other with x$format=6.
...	optional arguments
random.effects	logical. If TRUE, quantifies the overdispersion associated with the y-intercept of the data.

Value

a list of the slope and intercept of the best fit line as well as their standard errors and covariance.

Author(s)

Pieter Vermeesch and Mathieu Daëron

References

Daëron, M., 2023. Making the Case for Reconciled \Delta47 Calibrations Using Omnivariant Generalized Least-Squares Regression (No. EGU23-10066). Copernicus Meetings.

Daëron & Vermeesch, in prep. Omnivariant Generalized Least Squares Regression: Theory, Geochronological Applications, and Making the Case for Reconciled \Delta47 calibrations, Chemical Geology.

Examples

fn <- system.file('UW137.csv',package='IsoplotR')
UW137 <- read.data(fn,method='other',format=6)
fit <- ogls(UW137)

IsoplotR documentation built on Oct. 19, 2024, 5:07 p.m.