About slolay

slolay · ‎03-08-2012

Hi All I need to check that all the records in a earlier extracted dataset appear identically in the later extracted larger dataset. The first dataset has a unique recordid for each obs. I need to check that these recordids are in the second larger extract (of the same data) and then compare each variable value to see if any of the data values have changed. The second dataset is a later snapshot of the first and recordids could be missing, new ones added or data values changed. I just want to find out if my original data is missing or changed in anyway in the second dataset. Each obs has a unique recordid Any suggestions on how I can achieve this please? Many thanks Steve

slolay · ‎01-18-2012

Thanks for all your help guys I think SAS users could really benefit from log info regarding how many obs where read in for each input table as often where clauses are attached to them when reading in them in Have a good day Steve

slolay · ‎01-16-2012

Hi Thanks for the replies. Had a look at the SAS documetation: http://support.sas.com/documentation/cdl/en/sqlproc/62086/HTML/default/viewer.htm#a001360938.htm but nothing there about getting information about how many input records were read in. I hope SASkiwi is not right and it is not possible..it would be useful to have this information in the log. Regards Steve

slolay · ‎01-13-2012

Hi All Just started to use Proc SQL. Is it possible to get more informative messages in the log when running proc SQL code. It only seems to give the number of obs in the final output dataset but I would also like to see values for the input tables esp. when using joins and where clauses when reading the datasets in....still a DATA STEP man! Regards Steve

slolay · ‎12-19-2011

Thanks Guys!!!!

slolay · ‎12-16-2011

Hi I would like to produce a macro variable containing the start date and time a program was run. This will be used to add to the final report produced by the program I am currently using: /* create report start datetime macro variable */ data _null_; call symput('repstdttm',put(datetime(),datetime14.)); put "SAS program started: &repstdttm. "; run; this gives me 16DEC11:09:30 but I would like to get: Friday 16DEC2011:09:30 Is this possible using a format? I know there is the &sysday automatic variable but this is when the SAS session was started I believe. The way above should give me the time the datastep was executed Thanks in advance Steve

slolay · ‎12-12-2011

Thanks - does the job!...now to read up on it......

slolay · ‎12-09-2011

Hi I have a text variable containg dates in the format: YYYY-MMM-DD 2011-FEB-14 2002-SEP-22 etc I'm trying to convert these to SAS dates using input statement but get invalid data message for the INPUT statement Any help on how to get this working eg data test; date1='2011-AUG-12'; date2=input(date1,yymmdd11.); run; Thanks Steve

slolay · ‎12-05-2011

Thanks LexisC

slolay · ‎12-02-2011

Thanks KSharp

slolay · ‎12-02-2011

Hi Art How can I stop the PDV holding onto the last value in many-to-many merges? Using as example: data meds; input site subject conmed $10.; datalines; 1 1234 Med1 1 1234 Med2 ; proc sort;by site subject; run; data medhist; input site subject medhist $10.; datalines; 1 1234 Ulcer 1 1234 Wind 1 1234 Phlegm ; proc sort;by site subject; run; data all; merge meds medhist; by site subject; run; The required output is: 1 1234 Med1 Ulcer 1 1234 Med2 Wind 1 1234 Phlegm and not 1 1234 Med1 Ulcer 1 1234 Med2 Wind 1 1234 Med2 Phlegm I could use the lag function but thought there maybe another way to do the merge to avoid further programming to clean up afterwards Thanks Steve

slolay · ‎12-02-2011

Cheers Art, I think you've answered my questions, maybe, sort of, 😉 Will read the paper... Have a top weekend, farewell from Berlin Steve

slolay · ‎12-02-2011

Thanks Art From your reply I realise there is a way to get a cartesian join from data step coding(i just need to read the paper ;-)) but can you tell me if it's possible to 'switch' off the cartesian joining in PROC SQL (maybe it's also in the paper but a yes or no would help for now) Thanks Steve

slolay · ‎12-02-2011

Hi All How is it possible to do a cartesian product for a normal data step/ proc merge and how is it possible to 'switch off' the cartesain result from PROC SQL? Thanks Steve Using as example: data meds; input site subject conmed $10.; datalines; 1 1234 Med1 1 1234 Med2 ; proc sort;by site subject; run; data medhist; input site subject medhist $10.; datalines; 1 1234 Ulcer 1 1234 Wind ; proc sort;by site subject; run; data all; merge meds medhist; by site subject; run; proc sql; create table allsql as select a.*, b.medhist from meds as a left join medhist as b on a.site=b.site and a.subject=b.subject; quit;

slolay · ‎11-17-2011

Thanks Art Gonna go for Linlins option as a novel insight for me...and I don;t like arrays! :smileyshocked:

Online Status	Offline
Date Last Visited	‎03-22-2016 04:32 PM

Re: Are there online SAS Projects to Practice and Learn SAS?

Re: Are there online SAS Projects to Practice and Learn SAS?

Where to find online SAS projects to practice and learn SAS

Different Proc Report Output in Batch vs Interactive

Re: Shortcut to move to Command Line

Shortcut to move to Command Line

Re: Find out is a csv file is in PC or UNIX format?

Re: Find out is a csv file is in PC or UNIX format?

Find out is a csv file is in PC or UNIX format?

Re: PC SAS License needed for SAS/Connect to Unix

Re: Are there online SAS Projects to Practice and Learn SAS?

Problem using yymmdd11. informat

Assigning Missing Values Irrespective of Type

Where to find online SAS projects to practice and learn SAS

Proc Compare Datasets Which Increase In Size

Re: Proc SQL Log Messages

Proc SQL Log Messages

Proc SQL Log Messages

Full date from datetime value eg, Friday, 16DEC2011:09:30

Full date from datetime value eg, Friday, 16DEC2011:09:30

Problem using yymmdd11. informat

Problem using yymmdd11. informat

Cartesian ON and OFF

Cartesian ON and OFF

Cartesian ON and OFF

Cartesian ON and OFF

Cartesian ON and OFF

Cartesian ON and OFF

Re: How to Create a dataset with mutiple new Obs