ecoCopula (version 1.0.2)

predict.stackedsdm: Predictions from a stackedsdm object

Description

Predictions from a stackedsdm object

Usage

# S3 method for stackedsdm
predict(
  object,
  newdata = NULL,
  type = "link",
  se.fit = FALSE,
  na.action = na.pass,
  ...
)

Arguments

object

An object of class stackedsdm

newdata

Optionally, a data frame in which to look for variables with which to predict. If omitted, the covariates from the existing dataset are used.

type

The type of prediction required. This can be supplied as either a single character string, when is applied to all species, or a vector of character strings of the same length as ncol(object$y) specifying the type of predictions desired for each species. The exact type of prediction allowed depends precisely on the distribution, but for many there is at least "link" which is on the scale of the linear predictors, and "response" which is on the scale of the response variable. The values of this argument can be abbreviated.

se.fit

Logical switch indicating if standard errors are required.

na.action

Function determining what should be done with missing values in newdata. The default is to predict NA..

...

not used

Value

A list where the k-th element is the result of applying the predict method to the k-th fitted model in object$fits.

Author(s)

Francis K.C. Hui <[e-mail]>.

Details

This function simply applies a for loop, cycling through each fitted model from the stackedsdm object and then attempting to construct the relevant predictions by applying the relevant predict method. Please keep in mind no formatting is done to the predictions.

Examples

# NOT RUN {
X <- spider$x
abund <- spider$abund

# Example 1: Simple example
myfamily <- "negative.binomial"
# Fit models including all covariates are linear terms, but exclude for bare sand
fit0 <- stackedsdm(abund, formula_X = ~. -bare.sand, data = X, family = myfamily, ncores=2) 
predict(fit0, type = "response")

# }
# NOT RUN {
# Example 2: Funkier example where Species are assumed to have different distributions
abund[,1:3] <- (abund[,1:3]>0)*1 # First three columns for presence absence
myfamily <- c(rep(c("binomial"), 3),
       rep(c("negative.binomial"), 5),
       rep(c("tweedie"), 4)
       )
fit0 <- stackedsdm(abund, formula_X = ~ bare.sand, data = X, family = myfamily, ncores=2)
predict(fit0, type = "response")
# }