vegan (version 1.17-12)

scores: Get Species or Site Scores from an Ordination

Description

Function to access either species or site scores for specified axes in some ordination methods.

Usage

## S3 method for class 'default':
scores(x, choices, display=c("sites", "species"), ...)

Arguments

x

An ordination result.

choices

Ordination axes. If missing, default method returns all axes.

display

Partial match to access scores for sites or species.

...

Other parameters (unused).

Value

• The function returns a matrix of scores.

Details

Function scores is a generic method in vegan. Several vegan functions have their own scores methods with their own defaults and with some new arguments. This help page describes only the default method. For other methods, see, e.g., scores.cca, scores.rda, scores.decorana.

All vegan ordination functions should have a scores method which should be used to extract the scores instead of directly accessing them. Scaling and transformation of scores should also happen in the scores function. If the scores function is available, the results can be plotted using ordiplot, ordixyplot etc., and the ordination results can be compared in procrustes analysis. The scores.default function is used to extract scores from non-vegan ordination results. Most standard ordination methods of libraries mva, multiv and MASS do not have a specific class, and no specific method can be written for them. However, scores.default guesses where some commonly used functions keep their site scores and possible species scores.

If x is a matrix, scores.default returns the chosen columns of that matrix, ignoring whether species or sites were requested (do not regard this as a bug but as a feature, please). Currently the function seems to work at least for isoMDS, prcomp, princomp and some ade4 objects. It may work in other cases or fail mysteriously.

See Also

scores.cca, scores.decorana. These have somewhat different interface -- scores.cca in particular -- but all work with keywords display="sites" and return a matrix. However, they may also return a list of matrices, and some other scores methods will have quite different arguments.

Examples

data(varespec)
vare.pca <- prcomp(varespec)
scores(vare.pca, choices=c(1,2))