Comments by sulaif (Top 8 by date)

Thanks much for your response. Sorry for not being clear the first time. The dates cannot overlap even if the zip code is listed multiple times. In other words if there is a different zip code in between it should be grouping as such so the dates don't overlap while maintaining the date contiguity.

Uncompressed Dataset				
ID	ZIP	   START	        END	   Row
500	55106	7/8/09	     9/1/09	    1
500	55407	9/2/09	     11/23/09	2
500	55411	11/24/09     11/29/09	3
500	55407	11/30/09	 12/13/09	4
500	55429	12/14/09	 12/20/09	5
500	55407	12/21/09	 12/22/09	6
500	55407	12/23/09	 1/3/10	    7
500	55407	1/4/10	     1/10/10	8
500	55304	1/11/10	     1/13/10	9
				
Final Compressed Dataset 				
ID	ZIP	    START	    END	      Row
500	55106	7/8/09	    9/1/09	    1
500	55407	9/2/09	    11/23/09	2
500	55411	11/24/09	11/29/09	3
500	55407	11/30/09	12/13/09	4
500	55429	12/14/09	12/20/09	5
500	55407	12/21/09	1/10/10	    6
500	55304	1/11/10	    1/13/10	    7

You can see rows 6,7,8 are only grouped together for 55407 zip code with the START date being the MIN and END date being the MAX though it is repeated multiple times.