/* As ballarwd pointed out. Make a data step to represent your data,Nodoby would like to type it for you ! */ data have; input id name $ snack breakfast; cards; 1 net_amou 27.97 1 1 num_txns 1 2 2 net_amou 27.97 1 2 num_txns 1 2 ; data temp; set have; n+1; run; proc transpose data=temp out=temp2; by n id name; var snack breakfast; run; proc transpose data=temp2 out=want delimiter=_; by id; id _name_ name; var col1; run; ... View more

As always: store dates as dates and you can use simple where-statements. data have; length id $ 10 First_Qual_Month Txn_Month $ 6; input id First_Qual_Month Txn_Month; datalines; 1000361 Jul-19 Jul-19 1000361 Jul-19 Aug-19 1000361 Jul-19 Sep-19 100897 Jan-20 Oct-19 100897 Jan-20 Nov-19 100897 Jan-20 Dec-19 100897 Jan-20 Jan-20 100897 Jan-20 Feb-20 100897 Jan-20 Mar-20 ; data fixed_dates; set have; length FirstQualMonth TxnMonth 8; format FirstQualMonth TxnMonth mmyyd8.; FirstQualMonth = input(First_Qual_Month, monyy6.); TxnMonth = input(Txn_Month, monyy6.); drop First_Qual_Month Txn_Month; rename FirstQualMonth = First_Qual_Month TxnMonth = Txn_Month ; run; data want; set fixed_dates; where First_Qual_Month < Txn_Month; run; What should happen, if more than six obs have First_Qual_Month < Txn_Month ? ... View more