/* Calculate  Gini coefficient from a data set 

    Philip Dixon,  Mar 1992  */

/* read the data */
/*   details will need to change for other data formats */
/* the data step that does the calculation assumes that */
/*   the variable of interest is called X */


data ginix;
  retain npt 0;
  input x @@;			/* read a data value, possibly more than one */
  iboot = 1;			/*		on a line */
  npt = npt + 1;		/* update the number of data values */

  call symput("npt", npt);    
    /* store a macro variable with the number of points */
       
cards;
20 23 18 25 28 18
;
/* Calculate Gini coefficient for the original data */
 
proc sort data=ginix;     /* sort each bootstrap sample so that the */
  by x;			/* values are in increasing order.  This  */
				/* is required by the Dixon et al algorithm */
				/* to calculate the Gini coeff.  */
				/* (Ecology 1987)  */ 


data gini;
     
  set ginix end = last;			
  retain ginisum 0 		/* Accumulates (2i-n-1)*X  */ 
	sum 0			/* Accumulates X, so we can calculate mean */
	ipt 0;			/* index of data value, 1 for smallest, 
          				/*   npt for largest value. */

  npt =  symget('npt');      /* number of points in the data set*/
  
  ipt = ipt + 1;		/* Then add 1 to the number of data points */
  sum = sum + x;		/* Add the value to the sum */
  ginisum = ginisum + (2*ipt - (npt + 1))*x;	/* and add (2i-n-1)X to its sum */

  if last then do;               /* if this was the last observation */
    gini = ginisum /((npt-1)*sum);	/* calculate Gini coeff. for sample */
    mean =  sum/npt;
    output;			/* store on boots data set */
    end;

  keep npt gini mean;
  
proc print;
  title 'Gini coefficient';