Hi @monday89  Thank you for the prompt answer    I have updated my previous post with the code. Let me know! ... View more

Hi @monday89    Do you want to output a report or a dataset? In your code, you mention the RACE variable. Or it is not in the entry dataset. Is it a mistake? Thank you for the clarifications.   Based on the input data, you can try this: data demographics; input id 1 sex $ 3-8 ethnic $ 10-21 group $ 23-25; datalines; 1 male hispanic yes 2 male not hispanic yes 3 female hispanic yes 4 female hispanic no 5 female not hispanic no ; run; proc sort data=demographics; by group; run; ods output onewayfreqs=demographics_freq_sex; proc freq data=demographics; by group; tables sex; run; ods listing; ods output onewayfreqs=demographics_freq_ethnic; proc freq data=demographics; by group; tables ethnic; run; ods listing; data want; length _LABEL_ $20.; set demographics_freq_sex (rename=(sex=_LABEL_) drop=F_sex) demographics_freq_ethnic (rename=(ethnic=_LABEL_) drop=F_ethnic); run; proc print data=want; run; ... View more

Hi @radhikaa4  You can try this: proc sql; create table want as select a.City, a.Count, a.Count/b.total as Percent format=Percent8.1, a.firstDate format=MMDDYY10., a.LastDate format=MMDDYY10. from (select City, count(City) as Count, min(Dates) as FirstDate, max(Dates) as LastDate from table group by City) as a full join (select distinct City, count(*) as total from table) as b on a.City = b.City ; quit; You can apply a format (e.g. MMDDYY10.) directly after the calculation so that dates can be human-readable : e.g. min(Dates) as FirstDate format = MMDDYY10. To compute percent, I have merged you first query (alias 'a') with another another one (alias 'b') that retrieve the total number of records.   Alias "a" Alias "b" Final table combining alias "a" and "b"                 ... View more

Hi @monday89    If you want to create a table you can try the following code: data test; input ID 1 group $ 3-11 age 13-14 weight 16-18 gender $ 20-25 race $ 27-33 ethnicity $ 35-46 diabetes $ 48-51; datalines; 1 placebo 25 150 female white hispanic yes 2 treatment 35 123 male asian not hispanic no 3 placebo 23 153 female unknown hispanic no 4 placebo 34 124 male white hispanic yes 5 treatment 27 110 male white not hispanic yes 6 treatment 44 135 male asian not hispanic yes 7 treatment 23 100 female asian not hispanic no ; run; /* compute frequencies */ proc freq data=test; tables gender * group / out= group_gender (drop=percent); tables race * group / out= group_race (drop=percent); run; /* compute percentages */ proc sql; create table group_gender_n_pct as select a.gender,a.group,a.count,a.count/b.total as percent format=percent8.2 from group_gender as a full join (select group, sum(count) as total from group_gender group by group) as b on a.group = b.group order by gender; create table group_race_n_pct as select a.race,a.group,a.count,a.count/b.total as percent format=percent8.2 from group_race as a full join (select group, sum(count) as total from group_race group by group) as b on a.group = b.group order by race; quit; /* transpose data */ data group_gender_tr; length category $20.; merge group_gender_n_pct (where=(group="treatment") rename=(count = treatment_n percent=treatment_pct)) group_gender_n_pct (where=(group="placebo") rename=(count = placebo_n percent=placebo_pct)); by gender; category = gender; drop group gender; run; data group_race_tr; length category $20.; merge group_race_n_pct (where=(group="treatment") rename=(count = treatment_n percent=treatment_pct)) group_race_n_pct (where=(group="placebo") rename=(count = placebo_n percent=placebo_pct)); by race; category = race; drop group race; run; /* merge tables related to gender and race : table want */ data want; set group_gender_tr group_race_tr; run; proc print data=want; id category; var placebo_n placebo_pct treatment_n treatment_pct; run;   You can also create a table from PROC TABULATE as described by @mkeintz : proc format ; picture pct low-high='000.0%'; run; proc tabulate data=test out=have_count noseps; class group gender race; tables gender='00'x race='00'x , group=' '*(N='N'*f=comma6.0 colpctn='Pct'*f=pct7.0) /rts=10 ; run; data have_count_tr (drop= gender race _type_ _page_ _table_ PctN_100); format percent percent8.2; set have_count; length category $20.; if gender ne "" then category = gender; else category = race; percent = PctN_100 / 100; run; proc sort data=have_count_tr; by category; run; data want; merge have_count_tr (where=(group="treatment") rename=(n = treatment_n percent = treatment_pct)) have_count_tr (where=(group="placebo") rename=(n = placebo_n percent = placebo_pct)); by category; drop group; run; ... View more

Hi @monday89  It seems that you already posted this question: https://communities.sas.com/t5/SAS-Programming/proc-transpose-many-columns/m-p/611836 There are many answer attempts. Please check them.   This one for example should work. NB: you can't do that in a single proc transpose.   data demographics_freq; input category $1-6 group $8-16 frequency 18 percent 20-24; cards; female treatment 3 33.33 male treatment 6 66.67 female placebo 1 33.33 male placebo 2 66.67 ; run; data c.test; merge c.demographics_freq (where=(group="treatment") rename=(frequency = treatment_frequency percent = treatment_percent)) c.demographics_freq (where=(group="placebo") rename=(frequency = placebo_frequency percent = placebo_percent)); by category; drop group; run;   ... View more

Hi @SAS93    The SEQNO seems to be only an identification number.   In the BRFSS documentation, Adverse Childhood Experience is referenced as module 22. Responses to the 11 questions are respectively stored in variables ACEDEPRS, ACEDRINK, ACEDRUGS, ACEPRISN, ACEDIVRC, ACEPUNCH, ACEHURT1, ACESWEAR, ACETOUCH, ACETTHEM and ACEHVSEX (https://www.cdc.gov/brfss/questionnaires/pdf-ques/2019-BRFSS-Questionnaire-508.pdf)   However, all modules are not used every single year and in every single state (eg. https://www.cdc.gov/brfss/questionnaires/modules/category2018.htm)   What dataset do you have?   ... View more