R/mIFTO.parallel.invoke.gpxp.R

In AstroPathJHU/mIFTO: mIFTO: Multiplex Immunoflourescence Titration Optimization

#########################parallel.invoke.gpxp#################################

#'Used to loop through each image, call the gen func in parallel and return
#'Created By: Benjamin Green;
#'Last Edited 09/25/2019
#'
#'This function is desgined launch the parallelization of the gpxp function in
#' the mIFTO code. It is used to create a local environment with only necessary
#' data for the cluster objects.
#'
#'It is meant to be run through the RUN function
#'
#'
#' @param Concentration a numeric vector of concentrations used in the titration
#' @param x a unique identifier for the slide to be analyzed
#' @param y the numeric of which index from the concentration vector to use
#' @param Image.IDs the image coordinates of the current image as a comma
#' separated pair
#' @param Antibody_Opal the paired string for an antibody opal pair,
#' designated as "AB (Opal NNN)"
#' @param titration.type.name the type of titration that was performed
#' (TSA or Primary)
#' @param Thresholds a list of thresholds used for each concentration and slide
#' @param paths the paths to the data as a list, with an element for each
#'  concentration
#' @param connected.pixels the number of pixels that a pixel must be connected
#' to for positivity measures
#' @param flowout logical for whether or not flow like results will be produced
#' @param Opal1 the opal value of interest
#' @param decile.logical whether or not to run a decile approach analysis
#' @param threshold.logical whether or not to run a threshold approach analysis
#' @param cl cluster object
#' @return
#' @export
#'
#'
mIFTO.parallel.invoke.gpxp <- function (
    Concentration, x, y, Image.IDs, Antibody_Opal,
    titration.type.name, titration.type, Thresholds, paths,
    connected.pixels, flowout, Opal1,
    decile.logical, threshold.logical, cl) {
  err.val<-0
  tryCatch({
    #
    # define the environment for the cluster
    #
    my_env <- environment()
    parent.env(my_env) <- .GlobalEnv
    #
    # for each image gather the stats and return the images
    # to reduce RAM usage the code does this one image at a time
    # in addition parallel computing was implemented
    # to speed this up. Though the actual RAM usage is quite low
    # if I only carry the part of the image that is needed...
    #
    parallel::clusterExport(
      cl=cl, varlist=c("Concentration", "x", "y", "Antibody_Opal",
                       "titration.type.name","Thresholds","paths",
                       "connected.pixels","flowout","Opal1",
                       "decile.logical", "threshold.logical"),
      envir=my_env)
  }, warning=function(cond){
    return(list(err.val=cond))
  }, error=function(cond){
    return(list(err.val=cond))
  })

  #
  ###### need to add a try catch, but also need to determine what happens
  ###### when I throw an error instead of the envir
  tryCatch({
  small.tables.byimage<- lapply(
    Image.IDs[[x]][[y]],function(z) mIFTO::mIFTO.generate.pxp.image.data(
      Concentration, x, y, z, Antibody_Opal,
      titration.type.name, titration.type, Thresholds, paths,
      connected.pixels, flowout, Opal1,
      decile.logical, threshold.logical))
  }, warning=function(cond) {
    stop(cond)
    # if (typeof(pb.Object) != "character") {
    #   modal_out <- shinyalert::shinyalert(
    #     title = paste0('Error in small.tables for ',
    #                    x, ' 1to', Concentration[y]),
    #     text = paste0('Please check the computer resources, slide names, ',
    #                   'image layers correspond to protocol type, ',
    #                   'and that component data tiffs for ', x,
    #                   ' 1to',Concentration[[y]],' exist. Then contact ',
    #                   'Sigfredo Soto at ssotodi1@jh.edu for assistance.',
    #                   cond),
    #     type = 'error',
    #     showConfirmButton = TRUE
    #   )
    # }
  }, error=function(cond) {
    stop(cond)
    # if (typeof(pb.Object) != "character") {
    #   modal_out <- shinyalert::shinyalert(
    #     title = paste0('Error in small.tables for ',
    #                    x, ' 1to', Concentration[y]),
    #     text = paste0('Please check the computer resources, slide names, ',
    #                   'image layers correspond to protocol type, ',
    #                   'and that component data tiffs for ', x,
    #                   ' 1to',Concentration[[y]],' exist. Then contact ',
    #                   'Sigfredo Soto at ssotodi1@jh.edu for assistance.',
    #                   cond),
    #     type = 'error',
    #     showConfirmButton = TRUE
    #   )
    # }
  }, finally={})
  print("end of par")
  print(err.val)
  return(list(small.tables.byimage=small.tables.byimage, err.val=err.val))
}