I have data with this structure :
checkdate xclusive id1 id2 nstart nend
---------- -------- --- --- ------ ----
01/01/2003 F 1 1 96 120
01/01/2003 F 1 1 108 120
01/01/2003 F 1 1 108 120
01/01/2003 F 1 1 108 120
01/01/2003 F 2 1 96 108
01/01/2003 F 2 1 102 120
01/01/2003 F 2 1 102 120
01/02/2003 F 2 1 96 108
01/02/2003 F 2 1 96 132
01/02/2003 F 2 1 102 108
01/02/2003 F 2 1 102 120
01/02/2003 F 2 1 102 120
Start,End are in the range 1,186.
What I need is counts for distinct Date,Start,End (ids and xcl matter). There are overlaps in Start/End and I need to split them.
Above samples counts should be as :
checkdate xclusive id1 id2 nstart nend cnts
---------- -------- --- --- ------ ---- ----
01/01/2003 F 1 1 96 108 1
01/01/2003 F 1 1 108 120 4
01/01/2003 F 2 1 96 102 1
01/01/2003 F 2 1 102 108 3
01/01/2003 F 2 1 108 120 2
01/02/2003 F 2 1 96 102 2
01/02/2003 F 2 1 102 108 5
01/02/2003 F 2 1 108 120 3
01/02/2003 F 2 1 120 132 1
Any ideas on a fast routine to do this. TIA.
Cetin
