#!/usr/bin/gawk -f

BEGIN{
  seed=0;
  file="file";
  sample = 10;
  first = 1;
  notest = 1;
  noDatum = 0;
}

first{
  train="train" file;
  test="test" file;
  srand(seed);
  first = 0;
}

1{
  data[noDatum++]=$0;
}

END{
  testData = noDatum/sample;
  for ( i = 0; i < 10; i++ )
  {
    train = "train" i file;
    test = "test" i file;
    
    dataPrinted = 0;
    for ( d in data )
    {
      if ( dataPrinted >= i * testData && dataPrinted < (i+1)*testData ) print data[d] >> test;
      else print data[d] >> train;
      dataPrinted++;
    }
  }
}