tables {vegsoup} | R Documentation |
Method constancy
lists for each species the fraction (percentage) of samples in each partition (or the whole data set) the species occurs; contingency
lists just the number of occurrences. The latter is based on presence/absence of a species (layer-replicate), whereas method average
divides the sum of species abundances by the number of plots in which it occurs. A tabulation of partitioning(obj)
against a cross-classifying factor taken from a column of sites(obj)
can be obtained with the as.table
method.
## S4 method for signature 'Vegsoup' constancy(obj, percentage = TRUE, ...) ## S4 method for signature 'VegsoupPartition' constancy(obj, percentage = TRUE, ...) ## S4 method for signature 'Vegsoup' contingency(obj, ...) ## S4 method for signature 'VegsoupPartition' contingency(obj, ...) ## S4 method for signature 'VegsoupPartition' average(obj, ...) ## S4 method for signature 'VegsoupPartition' as.table(x, ...)
obj, x |
|
percentage |
|
... |
additional arguments passed to methods. For |
An importance table as returned by method average
lists for each species the average (mean) abundance of each species (stratum-replicate) in each partition.
Sensible results are only to be obtained if the object has either a semi-quantitative or quantitative cover scale defined (check is.ordinal(obj)
).
Note, if only presence/absence data is available average
returns the same results as [...], which bears no information as the summed abundance is just the number of occurrences.
See method quantile
for descriptive summaries of species abundances in partitions.
Specifying mode = "R"
in contingency
and average
allows for a kind of R-mode analysis where the grouping is defined by strata (layers
).
Note, this rarely makes sense if there is only one layer.
For multi layer (stratum) objects (length(layers(obj)) > 1
) a call to average
in R-mode implies that all species abundances in a layer are summed up and divided by layer contingencies (contingency(obj, mode = "R")
), in this case the number of species is found in a respective layer for each plot.
For constancy
, contingency
, and average
a object of class 'matrix'
with as many columns as there are partitions (strata if mode = "R"
) and as many rows as species (respectively plots). If mode is not specified, strata-replicates are treated separately.
For as.table
an object of class 'table'
with the levels of the cross-classifying factor in rows.
Roland Kaiser
const
and
importance
in labdsv,
quantile
data(barmstein) x <- VegsoupPartition(barmstein, k = 2) # constancy table for two partitions head(constancy(x, digits = 1)) # equivalent head(round(contingency(x) / as.vector(table(partitioning(x))) * 100, digits = 1)) # R-mode analysis # remove presence/absence standardisation decostand(x) <- NULL contingency(x, mode = "R") average(x, mode = "R") # tabulate partitioning vector againts sites column as.table(x, "expo") # tabulate partitiong vector againts other vector y <- rownames(subset(x, "Achnatherum")) # plot names with this species y <- y == rownames(x) # TRUE where the species occurs as.table(x, y)