dissimilarity: dissimilarity objects

Description Usage Arguments Value Examples

Description

creates a dissim_table which is required by bbgdm

Usage

1
2
3
4
5
6
7
8
dissim_table(sp.dat, env.dat, dism_metric = "number_non_shared",
  spline_type = "ispline", spline_df = 1, spline_knots = 2,
  coord.names = c("X", "Y"), intercept = FALSE, geo = FALSE,
  geo.type = "euclidean")

is.dissim_table(x)

is.spline(object)

Arguments

sp.dat

Presence absence matrix.

env.dat

Matrix of covariates for bbgdm; if using own data as covariates make sure to set sim_covar=FALSE, otherwise simulated covariates will be caluclated instead.

dism_metric

Dissimilarity to caluclate; If "bray_curtis" calculates Bray-Curtis dissimilarity, if "number_non_shared" returns number of shared species at site_ij which can be used in a binomial model for GDM.

spline_type

If "ispline" calculates ispline. If "bspline" calculates bc from spline package.

spline_df

degrees of freedom; one can specify df rather than knots. Default = 3.

spline_knots

The internal breakpoints that define the spline. Default = NULL for bs spline and 1 for ispline.

coord.names

c("X","Y") character.vector names of coordinates.

intercept

logical if TRUE it only returns the dissimilarities.

geo

logical If true geographic distance is calculated if

geo.type

type of geographic distance to estimate, can call 'euclidean','greater_circle' or 'least_cost'. If least_cost is called extra parameters are required (lc_data, minr and maxr).

x

dissim_table class

object

spline class object

Value

dissim_table dissimilarity table; creates a table of dissimilarities, and difference of covariates between each sites_ij. If select dissim="number_non_shared", will return "nonsharedspp_ij","sumspp_ij", which can be used as response variables in a binomial bbgdm.

Examples

1
2
3
x <- matrix(rbinom(100,1,.5),10,10) #toy presence absence matrix
y <- simulate_covariates(x,4)
diff_table <- dissim_table(x,y,dism_metric="number_non_shared")

skiptoniam/bbgdm documentation built on May 30, 2019, 1:05 a.m.