R/dr_maxnet.R

In bmaitner/pbsdm: Small Sample Size Species Distribution Modeling

#'Density-ratio SDM estimation with Maxnet
#'
#'dr_maxnet is an internal function for density-ratio estimation with Maxnet
#' @param presence_data dataframe of covariates
#' @param background_data dataframe of covariates
#' @param projection_data dataframe of covariates
#' @param formula Maxnet formula to use. Default (NULL) will use the Maxnet default. This parameter is called "f" in the maxnet function, but is renamed here as using "t" and "f" as object names is frowned upon.
#' @param regmult Maxnet regularization multiplier. Default is 1.
#' @param regfun Maxnet regularization function. Default is the Maxnet default.
#' @param addsamplestobackground If TRUE (the default), any presences that aren't in the background will be added.
#' @param clamp If TRUE (the default), predictions will be limited to ranges seen in the training dataset.
#' @param type Type of response required.  Defaults to link, exponential, cloglog, and logistic.
#' @param method one of either "fit" or "predict"
#' @param object fitted object returned by a dr_... function. Only needed when method = "predict"
#' @import maxnet
#' @importFrom utils getFromNamespace
#' @note The options f, regmult, regfun, and addSamplestobackground are only used when method == "predict",
#' the options clamp and type are only used when method == "predict". See the much better documentation for maxnet for more details.
#' @keywords internal
dr_maxnet <- function(presence_data = NULL,
                      background_data = NULL,
                      projection_data = NULL,
                      formula = NULL,
                      regmult = 1,
                      regfun = maxnet.default.regularization,
                      addsamplestobackground = TRUE,
                      clamp = TRUE,
                      verbose = FALSE,
                      method,
                      type = c("link", "exponential", "cloglog", "logistic"),
                      object = NULL){

  #Code to check inputs
  if(method=="fit" & (is.null(presence_data) | is.null(background_data) )){
    stop("When fitting maxnet, supply both presence and abscence data")

  }

  if(method=="predict" & (is.null(projection_data) )){
    stop("When predicting with maxnet, supply projection data")

  }


  #Code for fitting
  if(method == "fit"){

    #Generate presence/abscence vector

    pa_vector <- c(rep(1,nrow(presence_data)),
                   rep(0,nrow(background_data)))


    #Generate covariate data.frame

    pa_covariates <- as.data.frame(rbind(presence_data,
                                         background_data))

    # Specify formula
    if(is.null(formula)) {
      formula <- maxnet.formula(pa_vector, pa_covariates)
        }

    #Fit model
    ratio  <- maxnet(p = pa_vector,
                             data = pa_covariates,
                             f = formula,
                             regmult = regmult,
                             regfun = regfun,
                             addsamplestobackground = addsamplestobackground)


    model <- list(ratio = ratio,
                  method = "maxnet")

    class(model) <- "dr_estimate"
    return(model)

  }

  #Code for predicting

  if(method == "predict"){

    predict.maxnet <- getFromNamespace("predict.maxnet", "maxnet")

    prediction <- predict.maxnet(object = object$ratio,
                                 newdata = projection_data,
                                 clamp = clamp,
                                 type="exponential")

    return(prediction)
  }

}