As FreelanceReinh  said, the value at time=50 is SDF, SDF=1-CDF , it is  maybe just what you are looking for >50. proc lifetest data = Patients method=km plots=survival (cb = hw test atrisk) timelist=25 30 35 40 45 50; time Months*event_fl(0); title "Kaplan Meier Analysis on time to death"; label Months = "Time to Death"; run;     ... View more

What value would you expect changed_string to have if you ran:   changed_string='1'; changed_string='2'; changed_string='3'; Do you see what your problem is?     Or let's do it with numbers:   x=1; y=x+2; y=x+3; y=x+4; How should you modify those four statements so that y ends up with 10? x=1; y=x+2; y=y+3; y=y+4; Can you apply the same thing to your assignment statements?   ... View more

First useful and possibly quite critical comment:   Provide some example data in the form of a working data step posted in a text box. The data should be typical of what you have. The working data step is so we understand your data. For instance, with some "table" copied from a word processor or spreadsheet we cannot tell which of your variables are numeric or text, whether your "dates" are proper SAS date values or character. This a an extremely critical bit when doing any manipulation related to date manipulation as otherwise the first step is to create actual date values.   Second, make sure you are consistent internally in your description, use VARIABLE names (see first comment about data step). You start with " therapy will be defined as all eligible treatments ". Since there is no column or variable indicated as having a name of Treatment then we are left guessing what that actually means. Same with "Line of Therapy (LoT)", not clearly defined. Other instances of not defined, please fix them. There is nothing shown that identifies what a "CAR-T" is.   Third, since there may be some order dependence going on is a "Drug_A_B" actually different than a "Drug_B_A" value in later processing? ... View more

If space are important then include some spaces in the test data. data mixed_data; input mixed_variable $20.; datalines; aB#5dH0p2Lq8R!xJ7sE9 3kFy @6u1oPwQvZr4Xb2 s9TgY0o2I!qLw5mH8xN7 dC4pL!aB9eR2o7Xn1yH6 6fWvGm9@jP3tNqLr1iH2 bS5lE 7dR0oA1!xQ8zC3 Kp2Jr!L6yFw7XsNv9H0e 4uO9rH# tWq1E7sY2cL6 xZ3yF2!iQ8vS0l1Hr7nG 5pA1!m Hl3 v C6 ; It sounds like you are saying you want to eliminate everything except digits letters and spaces. data want; set mixed_data; want = compress(mixed_variable,' ','kad'); run; Which will generate: Obs mixed_variable want 1 aB#5dH0p2Lq8R!xJ7sE9 aB5dH0p2Lq8RxJ7sE9 2 3kFy @6u1oPwQvZr4Xb2 3kFy 6u1oPwQvZr4Xb2 3 s9TgY0o2I!qLw5mH8xN7 s9TgY0o2IqLw5mH8xN7 4 dC4pL!aB9eR2o7Xn1yH6 dC4pLaB9eR2o7Xn1yH6 5 6fWvGm9@jP3tNqLr1iH2 6fWvGm9jP3tNqLr1iH2 6 bS5lE 7dR0oA1!xQ8zC3 bS5lE 7dR0oA1xQ8zC3 7 Kp2Jr!L6yFw7XsNv9H0e Kp2JrL6yFw7XsNv9H0e 8 4uO9rH# tWq1E7sY2cL6 4uO9rH tWq1E7sY2cL6 9 xZ3yF2!iQ8vS0l1Hr7nG xZ3yF2iQ8vS0l1Hr7nG 10 5pA1!m Hl3 v C6 5pA1m Hl3 v C6 If you also want to eliminate spaces then do not put anything at all for the second argument. want = compress(mixed_variable,,'kad'); Or put some character you DO want to keep in the second argument. data want; set mixed_data; want = compress(mixed_variable,'A','kad'); run;   ... View more

I have renamed the flag-variables, if you insist on using bad names, you will have to edit the array statement. data work.have; input id $ flag_L1 $ flag_L2 $; datalines; 1 1L 0 2 1L 2L 3 1L 0 4 1L 2L 5 1L 2L ; data want; set have; length flag $ 4; array flags flag_L:; do i = 1 to dim(flags); if flags[i] ^= '0' then do; flag = flags[i]; output; end; end; drop i; run; ... View more

This accomplishes what you ask for the shown values. First is a data step to actually have values. data have; input cust_id date1 :date9. date2 :date9. date3 :Date9.; format date: date9.; datalines; 1001 15JAN2023 20FEB2023 30MAR2023 ; data want; set have; array d (*) date1 date2; do i= 1 to dim(d); date4=d[i]; output; end; format date4 date9.; drop i; run; Why? I don't see any practical use.   What do you intend to do with the resulting set? It may be this step just adds complications on the way to your goal. ... View more

Difficult to say without any explanation of the derivation of Date4  Anyway I do not see the relation of your last questions with the topic of this thread and if there is, why you don't follow the guidance as explained before Kindly ask you create a new topic to treat this new question pls ... View more

data CustomerOrders2; input CustomerID2 earlier_date OrderDate2 :date9.; format earlier_date OrderDate2 date9.; datalines; 1001 15dec2022 01JAN2023 1002 15dec2022 15FEB2023 1003 15dec2022 12DEC2023 1004 15dec2022 25JAN2024 ; run; i have the dataset like this CustomerID2  Earlier_date      OrderDate2    1001                     15dec2022       01JAN2023       1001                     15dec2022      15FEB2023        1001                     15dec2022      12DEC2023       1001                     15dec2022      25JAN2024          i want output dataset to be CustomerID2   Earlier_date    OrderDate2   Orderdateprev 1001                   15dec2022        01JAN2023     15dec2022  1001                   15dec2022        15FEB2023      01JAN2023 1001                   15dec2022        12DEC2023     15FEB2023   1001                   15dec2022        25JAN2024    12DEC2023   So i want output like above like the first orderdates orderdateprev will be earlier date the second order date's orderdateprev will be the first orderdate2 like that my flow will be  Could someone guide me here ... View more

How I would approach it: in a DATA step, create two datasets (ans1 and ans2), both containing id and new variable test loop through the test1 variable (COUNTW and SCAN functions), and write to ans1 if test is in A,B,C or to ans2 if not sort both dataset by id and test with NODUPKEY to remove duplicates in two final DATA steps, concatenate the tests (use BY, FIRST. and LAST., RETAIN, CATX) ... View more