R/utility.R

In ggcyto: Visualize Cytometry data with ggplot

#' Convert a list to a data.table
#' 
#' data.table version of ldply
#' @param .data a list
#' @return a data.table
#' @noRd 
.ldply <- function (.data,  ..., .id = NA) 
{
  index <- names(.data)
  if(is.null(index)){
    index <- seq_along(.data)
    .id <- NULL
  }
  

  res <- .do_loop(index = index, .data = .data, ..., .id = .id)
  res <- rbindlist(res)
  setkeyv(res, .id)
  res
}

#' @param index the index of the list, can be either character or numeric
#' @param  .fun the function to apply to each element of .data
#' @param ... other arguments passed to .fun
#' @param .id see help(ldply)
#' @noRd 
.do_loop <- function(index, .data, .fun = NULL, ..., .id = NA){

  lapply(index, function(i){
    
        dt <- .fun(.data[[i]], ...)
        dt <- as.data.table(dt)
        #append id
        if(!is.null(.id)){
          if (is.na(.id)) {
            .id <- ".id"
          }
          eval(substitute(dt[, newCol := i], list(newCol = .id)))
        }
        dt
      })
}
#' convert a flowSet to a data.table
#' @param .data flowSet
#' @noRd 
.fsdply <- function (.data, ..., .id = NA) 
{

  index <- sampleNames(.data)
  res <- .do_loop(index = index, .data = .data, ..., .id = .id)
  res <- rbindlist(res)
  setkeyv(res, .id)
  res
}

#' Generate a marginal gate.
#' 
#' It simply constructs an boundaryFilter that removes the marginal events.
#' It can be passed directly to ggcyto constructor. See the examples for details.
#' 
#' @param fs flowSet (not used.)
#' @param dims the channels involved
#' @param ... arguments passed to \link[flowCore]{boundaryFilter}
#' @return  an boundaryFilter
#' @examples 
#' data(GvHD)
#' fs <- GvHD[1]
#' chnls <- c("FSC-H", "SSC-H")
#' #before removign marginal events
#' summary(fs[, chnls])
#' 
#' # create merginal filter
#' g <- marginalFilter(fs, chnls)
#' g
#' 
#' #after remove marginal events
#' fs.clean <- Subset(fs, g)
#' summary(fs.clean[, chnls])
#' 
#' #pass the function directly to ggcyto
#' dataDir <- system.file("extdata",package="flowWorkspaceData")
#' gs <- load_gs(list.files(dataDir, pattern = "gs_manual",full = TRUE))
#' # with marginal events
#' ggcyto(gs, aes(x = CD4, y = CD8), subset = "CD3+") + geom_hex(bins = 64)
#'
#' # using marginalFilter to remove these events
#' ggcyto(gs, aes(x = CD4, y = CD8), subset = "CD3+", filter = marginalFilter) + geom_hex(bins = 64)
#' 
#' @export
marginalFilter <- function(fs, dims, ...){
  boundaryFilter(x = dims, ...)
}