occumb (version 1.2.1)

occumbData: Constructor for occumbData data class.

Description

occumbData() creates a data list compatible with the model fitting function occumb(). The element (i.e., covariate) names for spec_cov, site_cov, and repl_cov must all be unique. If y has a dimnames attribute, it is retained in the resulting occumbData object, and can be referenced in subsequent analyses.

Usage

occumbData(y, spec_cov = NULL, site_cov = NULL, repl_cov = NULL)

Value

An S4 object of the occumbData class.

Arguments

y

A 3-D array or a dataframe of sequence read counts (integer values). An array's dimensions are ordered by species, site, and replicate, and may have a dimnames attribute. A dataframe's columns are ordered by species, site, replicate, and sequence read counts. The data for missing replicates are represented by zero vectors. NAs are not allowed.

spec_cov

A named list of species covariates. Each covariate can be a vector of continuous (numeric or integer) or discrete (logical, factor, or character) variables whose length is dim(y)[1] (i.e., the number of species). The order of the species of the covariate values must correspond to that of the species dimension of y. NAs are not allowed.

site_cov

A named list of site covariates. Each covariate can be a vector of continuous (numeric or integer) or discrete (logical, factor, or character) variables whose length is dim(y)[1] (i.e., the number of sites). The order of the sites of the covariate values must correspond to that of the site dimension of y. NAs are not allowed.

repl_cov

A named list of replicate covariates. Each covariate can be a matrix of continuous (numeric or integer) or discrete (logical or character) variables with dimensions equal to dim(y)[2:3] (i.e., number of sites \(\times\) number of replicates). The order of the sites and replicates of the covariate values must correspond to that of the site and replicate dimensions of y. NAs are not allowed.

Examples

# Generate the smallest random dataset (2 species * 2 sites * 2 reps)
I <- 2 # Number of species
J <- 2 # Number of sites
K <- 2 # Number of replicates
data <- occumbData(
    y = array(sample.int(I * J * K), dim = c(I, J, K)),
    spec_cov = list(cov1 = rnorm(I)),
    site_cov = list(cov2 = rnorm(J), cov3 = factor(1:J)),
    repl_cov = list(cov4 = matrix(rnorm(J * K), J, K))
)

# A case for named y (with species and site names)
y_named <- array(sample.int(I * J * K), dim = c(I, J, K))
dimnames(y_named) <- list(c("common species", "uncommon species"),
                          c("good site", "bad site"), NULL)
data_named <- occumbData(
    y = y_named,
    spec_cov = list(cov1 = rnorm(I)),
    site_cov = list(cov2 = rnorm(J), cov3 = factor(1:J)),
    repl_cov = list(cov4 = matrix(rnorm(J * K), J, K))
)
# A real data example
data(fish_raw)
fish <- occumbData(
    y = fish_raw$y,
    spec_cov = list(mismatch = fish_raw$mismatch),
    site_cov = list(riverbank = fish_raw$riverbank)
)

# Get an overview of the datasets
summary(data)
summary(data_named)
summary(fish)