# Remove all columns from a PHYLIP-formatted dataset that include only 0's or that include any ?'s
# Make sure that the length of taxon labels is correct in the first line below
# Note that the code only copies the header as is; you have to manually revise the number of characters in the new file
# Andrew Hipp (ahipp@mortonarb.org), 4 July 2005

removeZeros <- function() {
  labelLength = 10
  NexusIn <- readLines(choose.files(default = "", caption = "Input PHYLIP-formatted RFLP data file", multi = FALSE), n = -1)
  NexusOut = new("character")
  NexusOut[1] = NexusIn[1]
  for (i in 2:length(NexusIn)) {
    NexusOut[i] = substr(NexusIn[i],1,labelLength) }
  numberOfTaxa = length(NexusIn) - 1
  for (i in (1 + labelLength):nchar(NexusIn[2])) {
    if (is.na(sum(as.integer(substr(NexusIn[2:length(NexusIn)],i,i))))) {print(paste("Found ambiguities in character",i - labelLength))}
    else if (sum(as.integer(substr(NexusIn[2:length(NexusIn)],i,i))) != 0) {
      print(paste("Included character",i - labelLength))
      for (j in 2:length(NexusIn)) {
        NexusOut[j] = paste(NexusOut[j], substr(NexusIn[j],i,i), sep = "") }}}
  writeLines(NexusOut, file.choose())
  return(NexusOut) }