0

I have a set of records(say 1000) residing in flat files(.txt). I need to extract them and put them in buckets. Each bucket has an upper cap of 300 records. Which means I'll need 4 buckets. I need to print them out to an excel sheet. Each time I hit the cap I'm suppose to print the bucket no.

eg:

  • Bucket 1

    • rec 1,
    • rec 2,
    • ...
    • rec 300

  • Bucket 2

    • rec 301,
    • rec 302,
    • ... -rec 600

  • Bucket 3

and so on. Whats the best most efficient way to do it? Here is what I am doing:

int lim = 300;

if(vector.size()>lim){
    n = Math.ceil((double) (vector.size() / 300.0));
}else{
    n=1;
}

//I'm meant to hard-code the bucket names
String[] name = {"Bucket1","Bucket2","Bucket3","Bucket4"};

for(int j=0; j<n;j++){
  buf.append("\n\t\t Name:"+name[0]);
  for(int i=0; i<size;i++){
    if (i > 0 && i < lim) {
      buf.append(",");
    }
    if (i < lim) {
      buf.append("\n\t\t\t\tP:" + vector.get(i));
    }           
  }
  size = size - lim;    
}
Ben
  • 13,297
  • 4
  • 47
  • 68
sloggers1894
  • 115
  • 2
  • 7
  • what is in buf and which object is it? are you trying to create a coma seperated value file? – Ben Oct 06 '11 at 17:35
  • Look into the [modulo operator](http://en.wikipedia.org/wiki/Modulo_operation). In Java, the operator is the percent symbol (%). – styfle Oct 06 '11 at 17:35
  • im sorry, but what good is a modulo here? it returns what is left after a integer division... – Ben Oct 06 '11 at 17:42
  • You have 4 buckets, so you can compute a hashcode of a line and make hashcode % 4 to get the index of the bucket in which this record should land. That is just simple hash partitioning. But won't take the 300 records into account. If you face that a bucket is already filled up, you can rehash or just randomly put it into another bucket. – Thomas Jungblut Oct 06 '11 at 18:07
  • I'm too old for homework...just simplified the actual problem. buf is nothing but StringBuffer. Yes I'm trying to create a csv file. – sloggers1894 Oct 06 '11 at 18:45

0 Answers0