uniquity2: Estimating biotic uniqueness of biological samples
In tobiasgf/uniquity: estimating uniquity of a site

Description Usage Arguments Value Author(s) Examples

This algorithm uniquity estimates the uniqueness of a set of communities

1
2
3

uniquity2(sstable, classes = NULL, weights = NULL, presabs = TRUE,
  background_sites = 50, rep = 50, non_correspondence = TRUE,
  total_coverage = 0)

`sstable`	a data.frame with with a species/site matrix (or an OTU table) that has sites/samples as columns and species/OTUs as rows, and observations as presence/absence data or abundance data.
`classes`	a list (dataframe) assigning samples to a set of unique classes (e.g. habitat-types). Each class represents an exclusive proportion af the investigated total area). (data.frame should contain these column names: site, class). If no classes are supplied, each samples is assigned to a unique class, and these will receive equal weights.
`weights`	a list (dataframe) with the weight (proportion of total) of each class. (data.frame should contain these column names: class, weight). If no weights are supplied, classes are assigned equal weights.
`presabs`	TRUE/FALSE. reduce species/site matrix to presence absence (0/1). Default is TRUE
`background_sites`	any positive number. The number of background sites to construct for counting the number of unique species/OTUs of the sample/site in focus (i.e the uniquity of the site/sample). each each compare. If `background_sites` < 1, the algorithm uses that proportion of the sites investigated as the number of background sites. If `background_sites` =>1, that absolute number of background sites wll be constructed. Default is 0.66 (=66% of samples/sites)
`rep`	any positive integer. The number of replicates pr site evaluation. (i.e. the number of times to construct a pool of background sites for evaluating each site/sample). Default is 100.
`non_correspondence`	TRUE/FALSE. If set to TRUE non-matching entries (classes, weights, sites) will be removed from the dataet. If set to FALSE, non non-matching entries will throw an error. Can be usefull for subsetting data, as subsetting will only be necessary in one part of the data (species-site table)
`total_coverage`	any number from 0 to 1. The fraction of total area investigated. If set to 0 the supplied weights will be used, unless the total weight exceeds 1, in wich case weights are scaled for a total of 1. If set to any number above 0 (including 1), weights are scaled for a total weight of `total_coverage`.

Function uniquity returns a list of results based on the input table and parameters.

simple_site_score the sum of inverse species weights, as a fast approximation of the estimated uniquity.
uniquity_score the simulated uniquity contribution of each site, relative to species richness.
adjusted_uniquity - number of curated (parent) OTUs
uniquity_table the final table with number of unique observations pr species per site
uniquity_species_scores sum of contribution per species.
Unique_species_avg average of contribution per species for sites where it occurs.
Summed_species_weights sum of weight pr species (used for drawing species for background sites).
Average_species_weight_pr_site average species weight pr site (with a weight).
Top20_unique_sites top 20 unique sites.
Top20_unique_species top 20 species with highest uniquity impact.
Accumulated_uniqueness uniquity uniquity metric per replicate.
site_richness species richness per site.
replicates The number of replicates used for each site uniquity estimation.
background_sites The number of background sites sampled.
Removed_classes classes removed due to non correspondence.
Removed_sites sites removed due to non correspondence.
total_coverage sum of site weights used in weighting of species.

Tobias Guldberg Frøslev

1 2	uniquity(my_table) uniquity(my_table, my_classes, my_weights, presabs = FALSE, rep = 50, size = 100, non_correspondence = TRUE, total_coverage = 0.5)