https://communities.sas.com/t5/SAS-Procedures/Looking-ahead-with-a-Data-Step/m-p/49491#M13442 I'm hoping someone can help me with the following:<BR /> <BR /> The following is my dataset:<BR /> <BR /> MKT RB_FLAG DATE1 DATE2<BR /> A RB 19SEP 17OCT<BR /> A NRB 19SEP 24OCT<BR /> A RB 19SEP 21NOV<BR /> B NRB 17OCT 17OCT<BR /> B RB 17OCT 24OCT<BR /> B RB 17OCT 14NOV<BR /> B NRB 17OCT 29DEC<BR /> B RB 17OCT 16JAN<BR /> B NRB 17OCT 23JAN<BR /> <BR /> For each MKT I need to take the first instance when RB_FLAG="RB" and the first instance after where RB_FLAG="NRB" then calculate the difference between DATE1 (for RB) and DATE2 (for NRB). Also, I'm only interested in the first occurence of RB within a MKT. Ultimately, I'm trying to arrive at the following table:<BR /> <BR /> MKT RB_FLAG DATE1 DATE2 WEEK_DIFF<BR /> A RB 19SEP 24OCT 5<BR /> B RB 17OCT 19DEC 10<BR /> <BR /> I have read some similar posts on "looking ahead" within a datastep but I'm struggling to adapt the suggestions to my problem.<BR /> <BR /> Thank you ahead of time for you help on this!<BR /> Rich Sun, 31 Jan 2010 01:22:03 GMT deleted_user 2010-01-31T01:22:03Z https://communities.sas.com/t5/SAS-Procedures/Looking-ahead-with-a-Data-Step/m-p/49491#M13442 I'm hoping someone can help me with the following:<BR /> <BR /> The following is my dataset:<BR /> <BR /> MKT RB_FLAG DATE1 DATE2<BR /> A RB 19SEP 17OCT<BR /> A NRB 19SEP 24OCT<BR /> A RB 19SEP 21NOV<BR /> B NRB 17OCT 17OCT<BR /> B RB 17OCT 24OCT<BR /> B RB 17OCT 14NOV<BR /> B NRB 17OCT 29DEC<BR /> B RB 17OCT 16JAN<BR /> B NRB 17OCT 23JAN<BR /> <BR /> For each MKT I need to take the first instance when RB_FLAG="RB" and the first instance after where RB_FLAG="NRB" then calculate the difference between DATE1 (for RB) and DATE2 (for NRB). Also, I'm only interested in the first occurence of RB within a MKT. Ultimately, I'm trying to arrive at the following table:<BR /> <BR /> MKT RB_FLAG DATE1 DATE2 WEEK_DIFF<BR /> A RB 19SEP 24OCT 5<BR /> B RB 17OCT 19DEC 10<BR /> <BR /> I have read some similar posts on "looking ahead" within a datastep but I'm struggling to adapt the suggestions to my problem.<BR /> <BR /> Thank you ahead of time for you help on this!<BR /> Rich Sun, 31 Jan 2010 01:22:03 GMT https://communities.sas.com/t5/SAS-Procedures/Looking-ahead-with-a-Data-Step/m-p/49491#M13442 deleted_user 2010-01-31T01:22:03Z https://communities.sas.com/t5/SAS-Procedures/Looking-ahead-with-a-Data-Step/m-p/49492#M13443 Investigate using BY GROUP processing within the DATA step and you will want to use RETAIN to track your DATE1 value across DATA step iterations.<BR /> <BR /> Scott Barry<BR /> SBBWorks, Inc.<BR /> <BR /> Recommended Google advanced search argument for this topic/post:<BR /> <BR /> data step by group processing site:sas.com Sun, 31 Jan 2010 02:46:04 GMT https://communities.sas.com/t5/SAS-Procedures/Looking-ahead-with-a-Data-Step/m-p/49492#M13443 sbb 2010-01-31T02:46:04Z