predict.logitreg: Posterior probability of analogue-ness for fossil samples

In analogue: Analogue and Weighted Averaging Methods for Palaeoecology

Description

Predict the posterior probability of analogue-ness between fossil and training set samples based on logistic regression fits.

Usage

## S3 method for class 'logitreg'
predict(object, newdata, group = "all", k = 1, ...)

Arguments

object	an object of class "logitreg", resulting from a call to logitreg.
newdata	an object of class "distance" representing the dissimilarity between fossil samples and those from the training set.
group	character; for which group(s) are predictions sought. "all", the default, indicates that predictions for all groups and the combined analysis are produced. Currently ignored.
k	numeric; the number of close modern analogues to each fossil sample to consider. Not currently used and may be removed from the method as varying k should require refitting the logistic regression model with that number of close modern analogues considered.
...	additional arguments passed to other methods.

Value

A matrix of posterior probabilities of analogue-ness is returned. The rows represent the fossil samples and the columns the groupings. There are g+1 columns where g is the number of groups. The g+1th column represents the Combined analysis which is the overall posterior probability that a fossil sample is an analogue to a training set sample (i.e. to any one of the groups).

Author(s)

Gavin L. Simpson

See Also

See logitreg for example usage. cma for extraction of close modern analogue information and analog for the underlying computations.

analogue documentation built on June 21, 2021, 1:08 a.m.