06-14-2016 11:28 PM
As shown in the figure below, I would like to merge Table 1 and Table 2 using common key IDs: card_nbr and transaction_name. However, I found out that the transaction names in table 2 are truncated.
|Table 1||Table 2|
|8000||Harvey Norman 12M: 0/12||8000||Harvey Norman 12M|
|7000||Ikea 36M: 0/12||7000||Ikea 36M|
How do I truncate the transaction name in Table 1 to match the data in Table 2? I have checked the rest of the data and only these installment transactions would have the symbol ":" followed by e.g: 0/12, 3/12. etc. I would like to remove anything from the symbol ":" onwards from Table 1.
Appreciate anyone's kind help to solve this coding issue.
06-14-2016 11:56 PM
Use operator EQT that tests equality up to the length of the shortest string:
data a; length t1 $8; do t1 = "aaba", "acdteg", "uret"; n + 1; output; end; run; data b; length t2 $8; do t2 = "aab", "acd", "uretav"; m + 10; output; end; run; proc sql; select a.*, b.* from a, b where t1 eqt t2; quit;
Need further help from the community? Please ask a new question.