Thanks, @Kurt_Bremser. I think this credit goes to you, as @urap considered as accepted solution. @urap , is it possible for you to take that off, please?
... View more
@Astounding: Thanks for correcting my comment for it and explained root of this. I really appreciate it. @Geo- Please consider this explanation, Ignore mine. Not intended to mislead you. 🙂
... View more
In this case, The explanation truly satisfied because you can execute code but gives you missing value as of you have not used w.d (3.0) which represent w= width and d =decimal. In our case, wid=3 ( because age may be 100 which is hard to see ) and no decimal. I hope it makes clear now. Try with w.d format and with out you will surly see difference.
... View more
data work.family;
infile datalines;
input @1 date_of_birth mmddyy10. @11 first_name $5. @17 age 3./* w.d informat */;
datalines;
01/05/1989 Frank 11
12/25/1987 June 13
01/05/1991 Sally 9
;
run; You are partially right. This explanation misleads you because of the following : Originally This program has two errors : 1) Error due to pointer at 15 2) Need to use w.d in format (Only this error is explained). I can say this explanation is not enough evidence for answer, however I can not consider it is wrong. It should include error 1) and 2) for complete explanation. Note: when I used to solve my base sas questions I found many of that questions, so I need to run and check those program to verify that answer and explanation. I think you should try that. I am sending you correct code as well.
... View more
For the Second kind of question, I suggest you Proc Format which is easy to use. Proc format; value $ change ( this is your choice of name) "number/of/word"="the number/of/ those words" ; run; For the first question, I am trying but not getting your question somehow, if you can explain I would like to help you out. I think you can use proc report for this kind of scenario but before that, you need to rearrange data.
... View more
Prof sql noprint; Create table want as select * From datasetA as x right join datasetB as y On x.id=y.id and x.price=y.price Quit; Note make sure price and ID should be same format either numeric or characters
... View more
I have imported this code in to SAS. I surely can say this is a comment/ explanation / annotation whoever write this program. Please have a look once in to attachment that I am sending with this. I hope it helpful. It gives you extra information that which variable you use as Denominator and Nominator for calculation.
... View more
No, Problem.It depends upon your choice. I usually try to work with DATA steps only (even if proc sql is good facility to do task). I use proc sql when I come across such as many to many merge or which data steps can't handle.
... View more
/** REFER THIS AS YOU ASKED "I need login and logout time of every users for every day"**/data x;
infile datalines;
input date $14. time $8. user $10.;
datalines;
18jun2018 9:00 user1
18jun2018 13:00 user1
18jun2018 10:40 user1
18jun2018 23:40 user1
20jun2018 13:40 user1
20jun2018 14:39 user1
18jun2018 9:00 user2
18jun2018 3:00 user2
19jun2018 1:40 user2
20jun2018 20:20 user2
18jun2018 3:20 user3
20jun2018 1:50 user3
18jun2018 19:30 user3
18jun2018 12:00 user3
;
run;
data x1;
set x;
new_time=input(time,hhmmss4.);
new_date=input(date,date9.);
run;
proc sort data=x1;
by user new_date ;
run;
data x2;
retain cnt;
set x1;
by user new_date;
if first.user then cnt=1;
else if first.new_date then cnt=cnt+1;
run;
proc sort data=x2;
by user new_date cnt;
run;
data x3;
format logon hhmm10.;
set x2;
retain logon ;
by user new_date cnt;
if first.cnt then logon=new_time;
rename time=logoff;
if last.cnt;
drop cnt new_date new_time;
run; please check out very last code which i updated( message no 9 reply.) this should work as per your requirement.
... View more
