About zqkal

zqkal · ‎05-30-2019

Thanks for your help. this solution works.

Kurt_Bremser · ‎05-18-2019

Transpose the ratio_: variables to a long dataset, sort it, and apply your logic to the sequence. Then the code is completely data-driven. See Maxim 33.

FreelanceReinh · ‎05-14-2019

Hi @zqkal, Just in case that you "would like to change the order of rows based on the values provided in the 'order' column ...," i.e., not create a new variable (new_desc), but just sort the observations in a certain way, you can possibly use an ORDER BY clause in PROC SQL. Example: proc sql; create table want as select * from have order by type, indexc(order, first(desc)), input(substr(desc,2),16.) desc; quit; This assumes that dataset HAVE is sorted by TYPE.

zqkal · ‎05-14-2019

Thanks for your help

novinosrin · ‎04-30-2019

%let values = A, C, D, Z; data want; do order=1 to countw("&values",','); values=scan("&values",order,','); output; end; run;

ballardw · ‎05-21-2018

Key is getting the numeric values from the character variable mo_freq; data want; input name $ group $ mo_freq $ value ; do month = input(scan(mo_freq,1),best.) to ( input(scan(mo_freq,2),best.)); output; end; datalines; wlf_01 actv 1-4 0 wlf_01 actv 5-7 1.15 wlf_01 actv 8-10 1.00 wlf_02 dflt 1-4 0 wlf_02 dflt 5-6 5.15 wlf_02 dflt 7-8 4.25 wlf_02 dflt 9-10 0 wlf_03 prog 1-10 2.23 ; run;

ballardw · ‎10-05-2017

Please describe you process in terms of the variables you show. It really is not clear how "next date doesn't have the group specified " in terms of grp1 and grp2 as there are two variables but you say "group specified" and since 2003 has 2 rows in the have data I am not sure what you mean by "2003 doesn't have row 2". If you describe things in terms of which values, such as the previous year group with grp1= xxx and grp2=yyy then the problem gets much clearer. But what if there were a 2002 set of records? Would you still be looking for the 2001? Or if the previous year group is 3 years earlier such as 2001 followed by 2004. Would you want data from 2003 (which may not exist) or 2002 (which also doesn't exist) or 2001?

zqkal · ‎09-24-2017

Thanks for your help

zqkal · ‎06-23-2017

Thank you for you help. This solution also works

Ksharp · ‎06-03-2017

Very interesting. data have; infile cards expandtabs truncover; input Type $ st_dt : $20. id ed_dt : $20. amnt; cards; AA 12/1/2016 0 1/1/2017 100.00 AA 1/1/2017 1 1/1/2017 200.00 AA 2/1/2017 2 1/1/2017 300.00 AA 3/1/2017 3 1/1/2017 10.00 BB 12/1/2016 0 2/1/2017 50.00 BB 1/1/2017 1 2/1/2017 105.00 BB 2/1/2017 2 2/1/2017 210.00 BB 3/1/2017 3 2/1/2017 140.00 CC 12/1/2016 0 3/1/2017 130.00 CC 1/1/2017 1 3/1/2017 124.00 CC 2/1/2017 2 3/1/2017 242.00 CC 3/1/2017 3 3/1/2017 298.00 ; run; data temp; set have(keep=type amnt); by type; n+first.type; if first.type then do; do i=1 to n; _amnt=.;output; end; _amnt=amnt;output; end; else do;_amnt=amnt;output;end; keep type _amnt; run; data want; ina=0; merge have(in=ina) temp; by type; if ina; run;

art297 · ‎05-12-2017

data DATASET1; input (V1 V2 V3) ($); cards; 01 AA 02 01 AA 08 01 AA 03 02 BB 24 02 BB 98 02 BB 24 03 CC 11 03 CC 99 03 CC 15 02 DD 54 ; data dataset2; set dataset1; retain v4; v4=max(v3,v4); run; Art, CEO, AnalystFinder.com

zqkal · ‎05-08-2017

This will creat the desired output. data need; set have; by id; total_FAS_1990_01 + FAS_1990_01; total_FAS_1991_01 + FAS_1991_01; total_FAS_1991_02 + FAS_1991_02; total_FAS_1991_03 + FAS_1991_03; total_FAS_1992_01 + FAS_1992_01; total_FAS_1992_02 + FAS_1992_02; total_FAS_1992_03 + FAS_1992_03; total_FAS_1993_01 + FAS_1993_01; run;

thomp7050 · ‎04-25-2017

data t1; input group $ amt ratio month total; datalines; A 1200 0.2 1 240 B 3290 0.1 2 0 C 2000 0.1 3 0 A 1000 0.4 1 240 B 5000 0.5 2 0 C 1430 0.6 3 0 ; run; data t2; set t1_NEW; FORMAT NEWTOTAL 8.2; retain NEWTOTAL; IF MONTH = 1 THEN DO; NEWTOTAL = TOTAL; END; if month GE 2 then do; NEWTOTAL = ratio * NEWTOTAL; end; run; Edit: Sorry, forgot the final run command!

LinusH · ‎10-08-2016

And given that, it seems that you have bottleneck in your environment. FULLSTIMER might give some more clues, but you probably need to use OS monitoring to get the full picture (basic OS measures are available in Environment Manager if you have that installed/configured).

Cynthia_sas · ‎11-26-2014

Hi: Did you look in your log? Were there any notes or messages about uninitialized variables? Where do you create the OBSNO variable? I do not see it in the COVER_PAGE creation program. My guess is that your test on ROWNUM isn't working because OBSNO doesn't exist. You could fix this easily, by putting OBSNO in the program where you create COVER_PAGE (Instead of dropping the loop var I). I would also expect that the misspelled "forntsize" will cause you a problem, too. You are also missing a "close" at the bottom of the code...should be ODS TAGSETS.EXCELXP CLOSE; cynthia

Online Status	Offline
Date Last Visited	‎05-30-2019 08:44 PM

Re: Determine which variable to use based on column values

Determine which variable to use based on column values

Create a dynamic logic based on data available

order observation based on the value provided in the column

Re: compare two numeric observations and retain the first instance

compare two numeric observations and retain the first instance

Loop through a list and create a dataset

How to Expand a dataset using value from one of the column

Look at previous values and create new observation

Re: SAS Merge two datasets with missing by variables

Re: compare two numeric observations and retain the first instance

Re: How to Expand a dataset using value from one of the column

Re: Macro to extract data from Oracle

Re: Macro to extract data from Oracle

Re: Macro to extract data from Oracle

Re: Determine which variable to use based on column values

Re: Create a dynamic logic based on data available

Re: order observation based on the value provided in the column

Re: compare two numeric observations and retain the first instance

Re: Loop through a list and create a dataset

Re: How to Expand a dataset using value from one of the column

Re: Look at previous values and create new observation

Re: SAS Merge two datasets with missing by variables

Re: Extract conditional information from text

Re: How to lag a column based on date difference

Re: How to get Max value of a column

Re: How to prefix array element to create dynamic variable

Re: How to lookup previous value of a raw to calculate new value

Re: How to read real time from SAS log file

Re: ODS Style format not working