Hi @Aexor    I have met the same problem several times,  so now I always use a combination of a table with distinct values and a macro loop.    /* Get distinct list of different values */ proc sql; create table mylist as select distinct xyz from large_data; quit; /* Macro to loop over list, get next value in list and do something based on the value */ /* Payload inside the %do loop could be any ny number of steps */ %macro loopmylist; /* loop over list */ %do i = 1 %to &sqlobs; /* Get next value from mylist and store it in macro variable */ data _null_; set mylist (firstobs=&i obs=&i); call symputx('thisvalue',xyz); run; /* Payload code - Do whatever should be done based on the actual value */ /* This is an example */ Proc sql; create table subset_&xyz as select * from large_data where xyz = "&thisvalue" /* quotes only in case of char values */ quit; /* End of Payload */ %end; %mend; %loopmylist;   ... View more

Hi @Ameurgen  Oh, I understand your answer to mean that you want the year shifted one month, so december is part of the winter season in the following year. You could easily obtain that with my code (change marked with red text😞   data want; set have; length Season $12; select; when (month(Date) in(12, 1, 2)) Season = 'Winter'; when (month(Date) in(3, 4, 5)) Season = 'Spring'; when (month(Date) in(6, 7, 8)) Season = 'Summer'; when (month(Date) in(9, 10, 11)) Season = 'Fall'; otherwise Season = 'n.a.'; end; Season = catx(' ', put(year(Date)+(month(date)=12),4.), Season); run;           ... View more

Hi @Ameurgen    You can make the Season variable as tekst like this:   data have; format date DATE9.; Date = '03oct2021'd; output; Date = '12dec2021'd; output; Date = '27jan2022'd; output; Date = '04may2022'd; output; Date = '18jul2022'd; output; run; data want; set have; length Season $12; select; when (month(Date) in(12, 1, 2)) Season = 'Winter'; when (month(Date) in(3, 4, 5)) Season = 'Spring'; when (month(Date) in(6, 7, 8)) Season = 'Summer'; when (month(Date) in(9, 10, 11)) Season = 'Fall'; otherwise Season = 'n.a.'; end; Season = catx(' ', put(year(Date),4.), Season); run; The result is:       ... View more

Hi @mmohotsi    The variable July in your input table is a SAS date with format date9. So all you need to create both macro variables is:     proc sql noprint; select min(July), max(July) into: Monthbegin, :Monthend from tmp2.date1; quit; Result:  45 %put &=Monthbegin &=Monthend; MONTHBEGIN= 22848 MONTHEND= 22872   ... View more

Hi @AmirSari  You say you want to create 100 views, but from your code it seems that the desired result is 20 views, each taking 5 pct of the full dataset.  To achieve this, an array is not necessary, You use the array do determine the number of iterations and the number of observations selected in each iteration, and this can be done simpler, so here is a slightly modified version of your code:   data have; do i = 1 to 10000; output; end; run; %let splitcount = 20; data _null_ ; lo = 1 ; increment = floor(divide(n, &splitcount)); do i = 1 to &splitcount; hi = floor(lo) + (increment-1); call execute (catx (" ", cats ("data v", put(i,8.)), "/", cats ("view=v",put(i,8.)), ";")) ; call execute (catx (" ", "set have (firstobs=", lo, "obs=", hi, ") ;")) ; call execute ("run ;") ; lo = floor(hi) + 1 ; end ; stop ; set have nobs = n ; run ; But I wonder how you you are going to use these views, as I cannot imagine any purpose that cannot be solved simpler with a sample from the full dataset or by adding a group variable similar to the view number. Then you can pull the wanted observations with the group number:   * Add group to data; data want; set have nobs=n end=eod; retain group 1; increment = floor(divide(n, &splitcount)); if _N_ > (increment * group) and not eod then group = group + 1; run; * Get data from single group (same as view v3); data d3; set want (drop=increment where=(group=3)); run; ... View more

Hi @kovamk    Yes. At least I never got it to work. I remember a Stackowerflow post that addressed the problem, and it said something about the code actually being there, but not working until F2 is pressed in every cell, so to me that equals "not working".   I found a paper with a receipe to use ODS / proc print with a tagset to cheate the excel table. I haven´t tried it, but it is worth exploring: https://www.pharmasug.org/proceedings/2014/CC/PharmaSUG-2014-CC07.pdf   ... View more

Hi @GN0001    Did you look at the file content? - It might be a korn script that executes a SAS program. If this is the case, the script contains a command to start SAS with the program path/name as one of the arguments. This gives you 3 possible ways of executing the SAS program:   Copy the program path/name and run it in a SAS session with %include. Run the script from a command line, if you have direct access to the server. Run the script in a SAS session by using the X command, if you have the XCMD option set.   ... View more

 @Alexander5    Please show the full program and the full log. This error is probably caused by a syntax error earlier in the program, but that is just guesswork until the whole program and log is available. ... View more

Hi @Yura2301    The problem is that you are using the wrong command. If the flow is submitted to Process Manager in SAS MC / Schedule Manager, the flow definition is stored in Process Manager's Flow Storage folder, and you can see it in the Flow Manager and run it from there. You can also run it from a command line using the Jtrigger command instead of Jrun, which requires an external flow definition file. https://www.ibm.com/docs/en/slpm/10.2.0?topic=commands-jtrigger   We use the Jtrigger command to run all flows (700+) in our daily batch, and all flows are scheduled with 'Manually to the scheduling server' as trigger, so they never run unless triggered by Jtrigger or mun manually from Flow manager. We have built a scheduler that establish flow order based om metadata registrations of jobs with input and output tables and trigger flows when all preceding flows in a chain have status Done.       ... View more

Hi @Frank_johannes    I would use a different approach. Instead of merging the data sets i think it is easier to extract ID's from all datasets and use statistics to reform output. Here is a working example that handles one to many input tables, I think the only requirements is that all tables contains a variable named ID, and this variable is of the same type in all datasets.    libname have 'c:\temp\testdata'; /* Test data */ data have.u10_clucose; do id = 2 to 20 by 7; output; end; run; data have.u28_tgly; do id = 1 to 5 by 2; output; end; run; data have.u28_toothdata; do id = 1 to 5 by 3; output; end; run; /* Collect information from all datasets */ proc contents data=have._all_ noprint short out=x; run; proc sql noprint; select memname into :memlist separated by ' ' from x; quit; data collect; length id &memlist 8.; stop; run; %macro m; %do i = 1 %to &sqlobs; %let thismem = %scan(&memlist,&i,%str( )); data tmp; set have.&thismem (keep=ID); &thismem = 1; run; data collect; set collect tmp; run; %end; %mend; %m; /* Process information to report */ proc sort data=collect; by id; run; proc means data=collect noprint max; output out=idlist (where=(_stat_='MAX') drop= _type_ _freq_); by id; var &memlist; run; ... View more

Hi @Anastasija98  sorry, hit the wrong button.   You aren't making things easy. It is a bit difficult to answer your question when you don't provide input data in a usable form and don't explain what you are trying to achieve. By reading to decipher your code, it seems that you want a boolean flag set for each numeric variable in your input, and set it to 1 if the absolute value is > 3 or else 0. IF this is the case (we cannot know, because we don't know all variables in input), there might be a simpler solution. The following code computes all the boolean variables and adds them to the output data set. Then you have them all in one place and can later select what you want for a given purpose.   data have; ID = '12345'; var_1 = 0.234; var_2 = -4; var_3 = -2.9; var_4 = 25; run; data _null_; set have (obs=1); array vlist _numeric_; length out_list $200; retain out_list; do i = 1 to dim(vlist); out_list = catx(' ', out_list, catt('out_',vname(vlist{i}))); end; call symputx('out_list', out_list); run; %put &=outline_list; data want (drop=i); set have; array vlist _numeric_; array outline_list &out_list; do i = 1 to dim(vlist); if abs(vlist{i}) > 3 then outline_list{i} = 1; else outline_list{i} = 0; end; run;     ... View more

Hi @Nietzsche    There might - as explained by others - be a use of this simple form of one-to-one reading in some special cases. I have written SAS programs since 1980, but never needed it. But it is important to know what's happening in a Data Step, and this guide exemplifies the basic behaviour of a data step including some tricky details, that might confuse a newbie, but it doesn't really explain what is going on.   Per default, a Data Step containing a SET statement performs a loop over input. When a SET statement is executed, the next observation is read into the Program Data Vector (PDV), and this continues until the last observation is read. At the end of each iteration, all variables in the PDV (with some exceptions) are written to output, and then the PDV is cleared. If the current observation is the last, the Step stops.   Now, what happens, if there are more than one SET statement? - The first SET statement reads all variables from the current input observation into the PDV. Then the next SET Statement adds all variables from the current input observation to the PDV, overwriting all values that might exist in the PDV with the same variable name. When the last observation in any of the input datasets is processed, the step stops.   So in youe example, you get Sex and Age from the Patients dataset, and ID, Height and Weight from the Measure dataset. The ID variable in the PDV is taken from Measure, because the ID already read into the PDV from Patients is overwritten. Since the ID's in the 2 input datasets are not aligned 1:1, the output is garbled, because Sex and Age comes from another ID.    The ID 9125 has correct values, because it happens to be observation 9 in both data sets, but 8125 is number 7 in the Measures dataset, but get Sex and Age from ID 4759. So the example code warns about things that could go really wrong. One might never notice the problem if there had been 11 observatiuons in both datasets. Note also, that there are only 9 observations in output, because one of the input data sets has only 9 observations, which causes the step to stop.    Common coding practice would be to use a MERGE statement instead of the two SET Statements. MERGE requires the input datasets to be sorted on the ID variable, which is used as a BY variable with the MERGE Statement in a Data Step. But there are cases where more than one SET statement is very useful, like using 2 SET Statements referring the SAME dataset, but with different starting point, to perform a look-ahead by getting values from the next observation to use in conjunction with the current observation. But then it is not a "blind" one-to-one reading as exemplified here, but controlled by different data set options, automatic end-variable and conditional execution of the second set statement.           ... View more