Skip to contents

Survey order and terminal groups of an ecan TWINSPAN classification

Source: R/clustOrder_twspn.R
clustOrder_twspn.Rd

Extract survey identifiers, their positions in the TWINSPAN dendrogram, and, optionally, the terminal groups already fitted by ecan::twinspan().

Usage

clustOrder_twspn(twin_ecan, cluster.group = TRUE, cluster.number = NULL)

Arguments

twin_ecan

An object returned by ecan::twinspan(). Both classical and modified TWINSPAN are supported, including species = FALSE.

cluster.group

Logical; include the terminal group, path and depth?

cluster.number

Optional positive integer. If supplied, it must equal the number of terminal groups actually present in twin_ecan.

Value

A data.frame sorted by Survey.order. With cluster.group = FALSE, columns are Survey and Survey.order. Otherwise columns are Survey, cluster (ecan group ID), Survey.order, path (binary division path), and depth (terminal division depth).

Details

This is the TWINSPAN counterpart of iPastoralist::clustOrder, the function defined in the original file cluster_survey_order.R.

The groups are taken from twin_ecan$classification$group, matched by survey identifier. They are never obtained by cutting a new clustering. To request a different classification, refit ecan::twinspan() with the desired n_clusters. That argument is a stopping target: the fitted number can be smaller when no group can be divided further.

Survey positions follow stats::as.hclust() for the supplied object. Within a terminal group, TWINSPAN has not resolved the order of individual surveys; ecan retains their input order and inserts zero-height graphical merges.

See also

first_ten_species_twspn(), plotClustAggregate_twspn()

Examples

veg <- rbind(S1 = c(9, 5, 0, 0), S2 = c(8, 4, 1, 0),
             S3 = c(0, 1, 5, 8), S4 = c(0, 0, 4, 7))
colnames(veg) <- c("Species A", "Species B", "Species C", "Species D")
tw <- ecan::twinspan(veg, modified = TRUE, n_clusters = 2,
                     min_size = 2, polish = "hill", species = FALSE)
clustOrder_twspn(tw)
#>   Survey cluster Survey.order path depth
#> 1     S3       1            1    0     1
#> 2     S4       1            2    0     1
#> 3     S1       2            3    1     1
#> 4     S2       2            4    1     1
clustOrder_twspn(tw, cluster.group = FALSE)
#>   Survey Survey.order
#> 1     S3            1
#> 2     S4            2
#> 3     S1            3
#> 4     S2            4