summary.mat {analogue}

Summarise Modern Analogue Technique models

Description

summary method for class "mat".

Usage

## S3 method for class 'mat'
summary(object, k = 10,
        digits = max(2, getOption("digits") - 4), ...)

Arguments

object

an object of class "cma", usually the result of a call to cma.

k

numeric; maximum modern analogues to use to summarise model fits.

digits

numeric; the number of significant digits with which to format results.

...

arguments passed to or from other methods.

Value

A list with the components below. The number of analogues used, k is returned as attribute "k".

summ

a data.frame containing the model fits for training set samples. See notes below.

tbl

matrix of summary statistics for an un-weighted model.

tbl.W

matrix of summary statistics for a weighted model.

call

the matched function call

quantiles

the quantiles of the distribution of pairwise dissimilarities for the training set, for "probs = c(0.01, 0.02, 0.05, 0.1, 0.2)".

Note

The returned component "summ" contains the following:

Obs:

the observed responses for the training set samples.

Est:

the fitted values of the response for training set samples based on the average of k-closest analogues.

Resi:

the residuals of the fitted model based on the average of k-closest analogues.

W.Est:

the fitted values of the response for training set samples based on the weighted average of k-closest analogues.

W.Resi:

the residuals of the fitted model based on the weighted average of k-closest analogues.

minDC:

dissimilarity of closest analogue in training set for each training set sample.

minResi:

smallest residual for an un-weighted model of size "k".

k:

size of model leading to minimal residual, "minResi".

minW.Resi:

smallest residual for a weighted model of size "k.W".

k.W:

size of model leading to minimal residual, "minW.Resi".

Author(s)

Gavin L. Simpson

See Also

mat, summary.

Examples

## Not run: 
## continue the RLGH example from ?join
example(join)

## fit the MAT model using the squared chord distance measure
swap.mat <- mat(swapdiat, swappH, method = "SQchord")
swap.mat

## model summary
summary(swap.mat)

## model summary - evaluating models using k = 1, ..., 20
## analogues instead of the default, 10.
summary(swap.mat, k = 20)

## End(Not run)