R/ds.exposome_scale_exposures.R

In isglobal-brge/dsExposomeClient: DataSHIELD client site Exposome functions

#' @title Scale ExposomeSet
#' 
#' @description Scales the exposures of an ExposomeSet object: centering by 
#' mean and scaling dividing by the standard deviation.
#'
#' @param Set \code{character} Name of the ExposomeSet to scale.
#' @param method \code{character} (default \code{"combined"}) If \code{"combined"} the scaling will be 
#' pooled (using the pooled mean and standard deviation), if \code{"split"} the scaling will be performed 
#' individually on each study server.
#' @param new.obj \code{character} (default \code{NULL}) Name that will take the scaled ExposomeSet. 
#' If \code{NULL} the name \code{"exposomeSet_scaled"} will be used.
#' @param datasources datasources a list of \code{\link{DSConnection-class}} objects obtained after login
#'
#' @return This function does not have an output. It creates a object on the study server named.
#' @export
#'
#' @examples
#' \dontrun{Refer to the package Vignette for examples.}

ds.exposome_scale_exposures <- function(Set, method = "combined", new.obj = NULL, datasources = NULL){
  
  if (is.null(datasources)) {
    datasources <- DSI::datashield.connections_find()
  }
  
  checkForExposomeSet(Set, datasources)
  
  if(is.null(new.obj)){
    new.obj <- "exposomeSet_scaled"
  }
  
  if(method == "combined"){
    # Get ExposomeSet meta-means and n
    cally <- paste0("exposome_scale_exposures_meansDS(", Set, ")")
    means <- DSI::datashield.aggregate(datasources, cally)
    # Calculate pooled means
    means_servers <- Reduce("rbind", lapply(means, function(x){
      x$means
    }))
    ns_servers <- Reduce("rbind", lapply(means, function(x){
      x$n
    }))
    ns_totals <- colSums(ns_servers)
    
    means_x_ns <- as.matrix(means_servers) * as.matrix(ns_servers)
    means_x_ns_total <- colSums(means_x_ns)
    
    pooled_means <- means_x_ns_total / ns_totals

    # Get pooled SDs
    ds.exposures_pData(set = Set, type = "exposures", name = "aux_exposures", 
                       exposures_type = "numeric", datasources = datasources)
    cols <- dsBaseClient::ds.colnames("aux_exposures")[[1]]
    variance <- unlist(lapply(cols, function(x){
      dsBaseClient::ds.var(paste0("aux_exposures$", x), type = "combined", datasources = datasources)$Global.Variance[, 1]
    }))
    pooled_sds <- sqrt(variance)
    
    # Create new scaled exposomeSets
    cally <- paste0("exposome_scale_exposuresDS(", Set, ", means = c(",
                    paste(unlist(pooled_means), collapse = ","),
                    "), sds = c(",  paste(unlist(pooled_sds), collapse = ","),"))")
    DSI::datashield.assign.expr(datasources, new.obj, cally)
    
  } else if (method == "split") {
    cally <- paste0("standardizeDS(", Set, ")")
    DSI::datashield.assign.expr(datasources, new.obj, cally)
  } else {
    stop("Invalid 'method' argument. Valid options are ['combined', 'split']")
  }
  
  if(method == "combined"){
    DSI::datashield.rm(datasources, "aux_exposures")
  }
  
}