Hi nice people, I am analyzing survival data using proc PHREG. The data has six strata. To prevent clutter and also show the difference in survival for continuous and categorical covariates graphically, I would like to only plot a single strata at two or three continuous covariate values or two or three categorical levels. I searched the community, but could not find such thing. I appreciate any help to take care of this. Thanks, ... View more

Hi all, Hope you all are doing great and stay safe during the pandemic. I would like to combine several "Select" statements where I calculate the average of the variables under one "Proc SQL". Below is the code I use and it works, but I use this code for each variable separately and producing unnecessary files. I want to get the average of other variables like length, age, ... for the same ID under this code. I appreciate your time and help. ... View more

Hi, I would like to merge two datasets by STRUCTURE_NUMBER_008 and Inspection_Year if available in both datasets and keep all the variables in the output. when I run the code, it gives zero observation from the merge statement.   data SNOW_FIRNSA05; merge all_data(in=one) snow_data(in=two) ; by STRUCTURE_NUMBER_008 Inspection_Year; if one; run; The format for the BY Variables is the same, but I do not know why the code does not work. I appreciate the time and help I anyone could help me fix this error. ... View more

Hi all, Here is the scenario from the attached dataset. The variables are STRUCTURE_ID, YEAR_BUILT, YEAR_RECONSTRUCTED, Inspection_Year, and CONDITION.  Each structure is inspected as shown in the Inspection_Year column. The data are sorted based on the STRUCTURE_ID and Inspection_Year. I would like to add another column Structure_Age based on the following conditions. If the structure is not reconstructed (YEAR_RECONSTRUCTED=0) or CONDITION data before reconstruction is not available then; Structure_Age = Inspection_Year - YEAR_BUILT If the structure is reconstructed, I want to see if the CONDITION is improved (higher CONDITION value than the year before  YEAR_RECONSTRUCTED in CONDITION column. then,  Structure_Age = Inspection_Year - YEAR_RECONSTRUCTED If the structure is reconstructed but the CONDITION is not improved (same CONDITION value as the year before YEAR_RECONSTRUCTED in CONDITION column. then,  Structure_Age = Inspection_Year - YEAR_BUILT data want; input STRUCTURE_ID $ YEAR_BUILT YEAR_RECONSTRUCTED INSPECTION YEAR CONDITION $ STRUCTURE_AGE; cards; /*STRUCTURE_AGE is 1983-1958 = 25 for CONDITION 5 below. */ 1 1958 0 1983 5 25; /*STRUCTURE_AGE is 1984-1958 = 26 for CONDITION 4 below. The rest of CONDITION 4 are 0, I dont need it. */ 1 1958 0 1984 4 26; 1 1958 0 1985 4 0; 1 1958 0 1986 4 0; 1 1958 0 1987 4 0; 1 1958 0 1988 4 0; /* The structure is reconstructed in 1992 and CONDITION is improved in 1993 inspection and it is different than 1988 CONDITION 4, therefore STRUCTIRE_AGE is 1993-1992 =1 as below. For the rest of CONDITION 8, STRUCTURE_AGE = 0, I dont need it. */ 1 1958 1992 1993 8 1; 1 1958 1992 1994 8 0; 1 1958 1992 1995 8 0; /* The CONDITION drops to 7 in 1996 and STRCUTURE_AGE= 1196-1992 = 4 as below. STRUCTURE_AGE for the rest of CONDITION 7 is 0, I dont need it. if the CONDIION changes to 6, I need STRUCTURE_AGE to be computed, if not to be 0. */ 1 1958 1992 1996 7 4; 1 1958 1992 1997 7 0; 1 1958 1992 1998 7 0; 1 1958 1992 1999 7 0; 1 1958 1992 2000 7 0; 1 1958 1992 2001 7 0; /* The CONDITION in 1983 is 6 and STRUTURE_AGE = 1983 - 1937 = 46 and for the rest of CODNTION 6 it is 0 as I don't need it. in 1990 the CONDIION is 5 and STRUCTURE_AGE 1990 - 1937 = 53 and for the rest of CONDITION 5 it is 0 as below. */ 10027 1937 0 1983 6 46; 10027 1937 0 1984 6 0; 10027 1937 0 1985 6 0; 10027 1937 0 1986 6 0; 10027 1937 0 1987 6 0; 10027 1937 0 1988 6 0; 10027 1937 0 1989 6 0; 10027 1937 0 1990 5 53; 10027 1937 0 1991 5 0; 10027 1937 0 1992 5 0; 10027 1937 0 1993 5 0; 10027 1937 0 1994 5 0; /* The bridge is reconstructed in 1994, but CONDITION is not improved in 1995, as it is the same CONDITION as 1994 inspection,therefore STRCUTURE_AGE is 0 as below.*/ 10027 1937 1994 1995 5 0;   I appreciate your time and help! ... View more

Hi all, I would like to remove a specific letter and the number just following it, but not a similar number from a character format. data have; A 606R1   3 A 607R2   1 A 607R3   1 A 607R1   1 A 607R     1   data want; A 606   3 A 607   1 A 607   1 A 607   1 A 607   1   The data set is large. The R stands for revision and the number just following it the revision number. I want to remove the R and the number following it (not replacing it with space). There might be numbers following the revision number which I want to keep. ... View more