I am very new to survival analysis. I am looking for differences between these two methods - Kaplan-Meier(K-M) vs. Cox Regression. KM Survival Analysis cannot use multiple predictors, whereas Cox Regression can. KM Survival Analysis can run only on a single binary predictor, whereas Cox Regression can run on both continuous and binary predictors. KM is a non-parametric procedure, whereas cox regression is a semi-parametric procedure. Please check the above points. If incomplete or incorrect, please suggest the changes.   ... View more

I am trying to create a variable in a new data set using a macro variable. I am trying to store names of numeric variable in a macro variable and then populate names in a variable in a new data set. I know it can be easily done with PROC CONTENTS. It's for my learning. I am not working on a real life project.   %macro test; proc sql noprint; select name into: nvar separated by " " from dictionary.columns where LIBNAME = "SASHELP" and MEMNAME = "CLASS" and type = "num"; quit;   %let varn=%sysfunc(countw(&nvar%str( ))); %DO i=1 %TO &varn.; %let varName = %qscan(%sysfunc(compbl(&nvar)),&i,%str( )); %put &varName;   data want; var1 = symget(&varName.); output; run; %end; %mend; %test;   Thanks in anticipation. Any help would be highly appreciated! ... View more

I have a dataset containing more than 5000 numeric variables. I need to rename all the numeric variables to var1- var5000 (Not exactly 5000 variables). I am labeling them as well so that one has mapping after renaming them. The problem is SAS can store max number of characters 65k. %macro numrename; *Getting numeric variable names proc sql noprint; select name into : nvars separated by " " from dictionary.columns where LIBNAME = "JS" and MEMNAME = "OUTWOE" and type = "num"; quit; *Count number of variables; %let varn=%sysfunc(countw(&nvars%str( ))); *Run for all the numeric variables; %DO i=1 %TO &varn; *Selecting Variable one by one; %let varName =%scan(%sysfunc(compbl(&nvars)),&i,%str( )); %let vara = var&i.; *Renaming variables data outputt; set out_woe (rename = (&varName. = &vara.)); label &vara. = "&varName."; run; %mend; ... View more

I have written 12k lines of code over 2 years. I want some portion of code to be run for trial period (4 months). After that it should stop working unless the original developer of the macro wants to make it work.     What i could think off creating a reference variable that calculates difference of today's date and 2nd quarter of 2016 - x = intck('month', date(),yyq(floor(exp(8))- 964, ceil(arcos(0.5)))); If x <= 0 then stop macro.  It's too easy for intermediate SAS user to crack the above code. INTCK, date() functions are easy to understand. Is there any geek way to do it? Any trick with DO LOOP / SQL to create a password that works only till 2nd quarter of 2016? Any help would be highly appreciated! ... View more

I am working on a business problem that focuses on targeting credit card customers whom bank should increase credit card limit.  Has anyone worked on the similar model? What kind of techniques and data sets can be used for building this model? Can we predict credit card utilization to make it closer to the objective? ... View more

I want the following PROC SQL code to be run for multiple variables. Till now, it is designed for a single variable i.e. Balance. I want this to be run for multiple variables in a loop. The code calculates ratio and total sum of balance for a time period.   Data Sets used in the code -    data example1; input ID Months Revenue Balance; cards; 101 1 3 90 101 2 33 68 101 3 22 51 101 4 3 90 101 5 33 65 101 6 22 54 101 7 3 92 101 8 33 65 101 9 22 55 101 10 3 96 101 11 33 65 101 12 22 54 102 1 100 18 102 2 58 62 102 3 95 97 102 4 100 18 102 5 58 65 102 6 95 92 102 7 100 11 102 8 58 62 102 9 95 92 102 10 100 15 102 11 58 60 102 12 95 91 ; data trend; input x y; cards; 1 3 4 6 1 6 7 9 1 9 10 12 1 12 ; run;     proc sql noprint; select cat("sum(case when Months =",y," then Balance else . end)", "/ sum(case when Months =",x," then Balance else . end) as Balance_",y,"_",x,"_ratio") into :ratio separated by "," from trend; select cat("sum(case when ", x,"<= Months <=",y, " then Balance else . end) as Balance_",x,"_",y,"_Sum") into :loop separated by "," from trend; create table Output_trend as select ID, &ratio., &loop.,  from example1 group by ID; quit;   For example : Variables are specified in the following "vars" macro paramer and then run the above PROC SQL code for the variables mentioned in the macro variable.   %let vars = Balance Revenue; %let n=%sysfunc(countw(&vars,%str( )));   ... View more

 data example1; input ID Months Revenue Balance; cards; 101 1 3 90 101 2 33 68 101 3 22 51 102 1 100 18 102 5 58 6 102 3 95 97 103 1 47 94 103 2 85 64 103 4 9 88 ;   I am trying to automate the following steps - sum(case when Months = 1 then Balance else 0 end) as Balance_1, sum(case when Months = 2 then Balance else 0 end) as Balance_2, sum(case when Months = 3 then Balance else 0 end) as Balance_3, sum(case when Months = 4 then Balance else 0 end) as Balance_4, sum(case when Months = 5 then Balance else 0 end) as Balance_5   My code : proc sql noprint; select distinct "sum(case when Months = "|| put(Months,z2.0) ||" then Balance else 0 end) as Balance_"|| put(Months,z2.0) into :loop separated by "," from example1; create table Output as select ID, &loop. from example1 group by ID; quit;   It works fine. The only problem with the code is it can only be run after converting the variable to character with PUT function. If i remove PUT function, the code does not work. I want the code to be run without converting the variable to character operand. ... View more