ds.rNorm<-function(samp.size=1,mean=0,sd=1, newobj="newObject", datasources=NULL){
  
toAssign<-paste0("rNormDS(",samp.size,",",mean, ",", sd, ")")


  # if no opal login details are provided look for 'opal' objects in the environment
  if(is.null(datasources)){
    datasources <- findLoginObjects()
  }
  
  if(is.null(toAssign)){
    stop("Please give the name of object to assign or an expression to evaluate and assign.!\n", call.=FALSE)
  }
  
  # now do the business 

  # Because this is DataSHIELD, the simulated values in two different
  # studies could be the same because when no seed is set, "one is created
  # from the current time and the process ID". As the time in each
  # server may be very similar, you rely on process ID to make different
  # servers different. Certainly working on local VMs this often leads
  # to the same set of random numbers being generated. So, the ds.rNorm
  # function generates n=samp.size random vectors for n=samp.size studies
  # and the second call to ds.assign (below) picks the kth vector for the
  # for the kth study. 

#Randomly simulate n=samp.size vectors
  for(j in 1:length(datasources))
  {
  newobj.s<-paste0(newobj,j)
  datashield.assign(datasources, newobj.s, as.symbol(toAssign))
  }

#Select the values in the kth random vector for the kth study
  for(k in 1:length(datasources))
  {
  newobj.s<-paste0(newobj,k)
  datashield.assign(datasources[k], newobj, as.symbol(newobj.s))
  }

  # check that the new object has been created and display a message accordingly
  finalcheck <- isAssigned(datasources, newobj)
  
}