RAM (version 1.2.1.3)

dissim: Calculate Dissimilarity Matrix Data

Description

These functions calculate different measures related to dissimilarity matrices. All of these functions allow you to specify one of many dissimilarity indices to be used.

Usage

dissim.clust(elem, is.OTU=TRUE, stand.method=NULL, dist.method="morisita", clust.method="average")
dissim.eig(elem, is.OTU=TRUE, stand.method=NULL, dist.method="morisita")
dissim.ord(elem, is.OTU=TRUE, stand.method=NULL, dist.method="morisita", k=NULL)
dissim.GOF(elem, is.OTU=TRUE, stand.method=NULL, dist.method="morisita")
dissim.tree(elem, is.OTU=TRUE, stand.method=NULL, dist.method="morisita", clust.method="average")
dissim.pvar(elem, is.OTU=TRUE, stand.method=NULL, dist.method="morisita")

Arguments

elem

an ecology data set that can be an OTU table or a taxonomy abundance table. See RAM.input.formatting for details.

is.OTU

logical, whether the ecology data sets are OTU tables or taxonomy abundance matrices. See RAM.input.formatting for details.

stand.method

optional, if is.null, the standardization method for data transforamtion; must be one of the following: "total", "max", "frequency", "normalize", "range", "standardize", "pa", "chi.square", "hellinger", "log". See also decostand.

dist.method

the dissimilarity index to be used; one of "manhattan", "euclidean", "canberra", "bray", "kulczynski", "jaccard", "gower", "altGower","morisita", "horn", "mountford", "raup", "binomial", "chao", or "cao". See also vegdist.

k

the number of dimensions desired. If NULL, the maximum value will be calculated and used.

clust.method

the method used for clustering the data. Must be one of "ward", "single", "complete", "average", "mcquitty", "median", or "centroid". See also hclust.

Value

dissim.clust

returns a hierarchical clustering of the dissimilarity matrix.

dist.eigenval

returns the eigenvalues of the dissimilarity matrix.

dissim.ord

returns a list: the first item is the the ordination distances, the second is the dissimilarity matrix distances.

dissim.GOF

returns the goodness of fit values of the dissimilarity matrix, for various numbers of dimensions used.

dissim.tree

returns a list: the first item is the tree distances, the second is the dissimilarity matrix distances.

dissim.pvar

returns a numeric vector containing the percent variation explained by each axis (where each sample corresponds to an axis).

See Also

decostand, vegdist, hclust, dissim.plot

Examples

data(ITS1)
# calculate clustering, using default method
dissim.clust(ITS1)
# calculate tree distances, specifying a distance method
# (but use default clustering method)
dissim.tree(ITS1, dist.method="euclidean")
# calcualte ordination distances, specifying both distance
# and ordination methods
dissim.ord(ITS1, dist.method="bray", k=3)