# R x64 3.2.3 written by Ali Khalighifar and Laura Jiménez on June, 12, 2017
# The last edit was on Oct, 26, 2017 by Ali Khalighifar
# This script generates the species population using lognormal distribution in suitable format of EstimateS version 9.

### Generates sampling days, 100 real species, 10 error species, and 100 replicates 
binsize=c(5,7,10,15,20,25,50,75,100,125,150,175,200,300,400,500,600,700,800,900,1000)
nsp <- 100  
nerr <- 10 
nsim <- 100  

# Select the parameters of the lognormal distribution (mean, standard deviation, error abundance value, etc.)
meanl <- 1.5 
sdl <- 2 
lam <- rlnorm(nsp,meanlog=meanl,sdlog=sdl)
per.ln <- 0.000001 
lam.error <- qlnorm(per.ln, meanlog =meanl, sdlog =sdl)
lams <- c(lam,rep(lam.error,nerr))

##loop over each binsize (sampling days) and export a text file for each one
setwd("working Directory")
loop=0
for(k in binsize){
  input_file <- file(paste0(paste(nsp+nerr,"sp",per.ln, "err", nsim,"sim",k,"days",sep="_"),".txt"), open = "at")
  loop=loop+1
  for(j in 1:nsim){
    repmatrix=matrix(nrow=(nsp+nerr),ncol=k)
    for(i in 1:(nsp+nerr))
      # abundance matrix
      repmatrix[i,]=rpois(k,lams[i])
    # generate incidence matrix
    repmatrix <- (repmatrix>0)*1
    if(j==1){
      # Design the format of input files for EstimateS
      cat("*MultipleSampleSets*\t", nsim, "\t", paste0(k, "_days"), "\n", file = input_file, sep = "")}
    
    cat(paste("rep", j, sep = " "), "\t*SampleSet*\t", "1\t", "0\t", "0\n", file = input_file, sep = "")
    cat(nsp+nerr, "\t", k, "\n", file = input_file, sep = "")
    write.table(repmatrix,file=input_file, sep = "\t",
                row.names=F, col.names=F, append = T)
    
  }
  flush(input_file)
  close(input_file)
}