Ksharp Tracker

Re: Weird SAS error messages on quoted strings and white spaces

Ksharp — Fri, 17 Jul 2026 06:53:29 GMT

To suppress this Note or Warning, try this option:

option noquotelenmax;

Re: Alphabetical order of levels in proc gradar

Ksharp — Thu, 16 Jul 2026 06:20:14 GMT

Tom,
I pad some blanks before values to customize its order.
Sine the blank is before any A-Z and printable characters in ASCII table, if you order these value, the value which has top blanks is the first one.
For example:
' A' would be before 'A' or any other printable characters.

So I use "ORDERACROSS=FORMATTED" to let sas order its formatted value to get the order I need .
But sometimes I think "ORDERACROSS=INTERNAL" would be that one, but after testing it ,only FORMATTED get job done .

Re: Alphabetical order of levels in proc gradar

Ksharp — Wed, 15 Jul 2026 07:09:45 GMT

proc format;
value DM_range    1 = '20-25' 2 = '25-30' 3 = '30-35' ;
 
data wide;
input DM_range  a_flavor b_flavor c_flavor d_flavor e_flavor f_flavor g_flavor  h_flavor;
 
 
format DM_range DM_range.;
output;
 
datalines;
1 9.43  6.60  37.74 21.70 0.00  0.00 0.00 27.36
2 10.73 20.49 24.88 25.85 5.00  0.00 0.00 20.98
3 17.56 24.43 35.88 38.17 18.58 3.54 4.42 4.58
;
run;
 
proc sort data=wide out=long;
    by DM_range ;
run;
 
/* Transpose texture */
proc transpose data=long name=flavor_label out=long1(rename=(COL1 = flavor_value) );
    by DM_range;
    var a_flavor b_flavor c_flavor d_flavor e_flavor f_flavor g_flavor h_flavor;
run;  
 
  PROC PRINT DATA=long1 ;
RUN ;
 
 
/*rename levels if variable texture*/
data longnew;
length flavor_label $ 20;
    set long1;
    if flavor_label = 'a_flavor' then flavor_label = '       Astringent';
    else if flavor_label = 'b_flavor' then flavor_label = ' Oily';
else if flavor_label = 'c_flavor' then flavor_label = '     Green/Grassy';
else if flavor_label = 'd_flavor' then flavor_label = '  Nutty';
else if flavor_label = 'e_flavor' then flavor_label = '    Herbaceus';
else if flavor_label = 'f_flavor' then flavor_label = 'Spicy';
else if flavor_label = 'g_flavor' then flavor_label = '      Floral';
else if flavor_label = 'h_flavor' then flavor_label = '   None';
run;
 
PROC PRINT DATA=longnew ;
RUN ;
 
 
 /* Set the graphics environment */
goptions reset=all cback=white border htitle=12pt htext=12pt;
footnote1  j=c h=10pt "N = 442";  
title1 j=c h=13pt 'Comparison of flavor by DM range for BL516 (relative %)';
 proc gradar data=longnew;
   chart flavor_label / freq=flavor_value  ORDERACROSS=FORMATTED
                 overlay=DM_range
starcircles=(25 50)
                 cstarcircles=black
                 cstars=(aqua red green)
                 wstars=(2 2 2)
                 cframe=white
                 ;
run;
quit;

Re: Sum values over repeated variable values

Ksharp — Sat, 11 Jul 2026 07:27:25 GMT

As Tom said, you need to transform these characters into numbers and calculated/sum it again and concatenate them back .


data mytab;
    length 
           Value $15 
           VAR1 $80 
           VAR2 $50 
           VAR3 $50;
    
    infile datalines delimiter='|' missover dsd truncover;
    input  Value $ VAR1 $ VAR2 $ VAR3 $;
datalines;
   8 ( 20.5%)|Overall (*)||
   4 ( 10.3%)|ClassA|All|
   1 (  2.6%)||A1|A1
   1 (  2.6%)||A1|A1
   1 (  2.6%)||A1|A1
   1 (  2.6%)||A1|A1
   1 (  2.6%)||A1|A1
   1 (  2.6%)||A2|A2
   1 (  2.6%)|Class345|All|
   1 (  2.6%)||NF1|NF1
;
run;

data mytab2;
 set mytab;
 part1=input(scan(value,1,'( )'),best.);
 part2=input(scan(value,-1,'( )'),percent.);
run;
proc summary data=mytab2 ;
 by var1-var3 notsorted;
 var part1 part2;
 output out=mytab3(drop=_:) sum=;
run;
data mytab4;
 length value $ 20;
 set mytab3;
 value=catx(' ',part1,'(')||tranwrd(put(part2,percent7.1 -r),'  ',' ')||')';
 drop part1 part2;
run;

proc print data=mytab4 noobs;
run;

Re: proc tabulate with 2 nested variables for columns, 1 sorted ascending and 1 sorted descending

Ksharp — Fri, 10 Jul 2026 08:34:08 GMT

As ballardw said, it would be better if you could post some data and the desired output report.
So we can test your code and get output .

Re: Accuracy, format and round

Ksharp — Thu, 09 Jul 2026 07:35:12 GMT

For your first question, sas is unable to store 73.75 exactly in computer sometimes, so you can't always expect the expression is equal. If you want to compare them ,the best way is using ROUND() function.

if you need to display "73.7499999999999999999999999999999999" , you need to resort to PICTURE format.

proc format;
picture fmt
low-high='09.99999999999999'
;
run;

data test;
	x1=(36.9-31)*12.5;
	x2=73.75;
	if x1 <x2 then flag="Y";
put x1=  fmt32.20 x2= fmt32.20  flag=;

    call missing(flag);
	if round(x1,0.1) <round(x2,0.1) then flag="Y";
put x1=  32.20 x2= 32.20 flag=;



    z1=round(x1,0.1);
	z2=round(x2,0.1);
put z1=  32.20 z2= 32.20;

    w1=strip(put(x1,5.1));
	w2=strip(put(x2,5.1));
put w1= w2=;
run;

x1=73.74999999999999 x2=73.75000000000000 flag=Y
x1=73.75000000000000000000 x2=73.75000000000000000000 flag=
z1=73.80000000000000000000 z2=73.80000000000000000000
w1=73.7 w2=73.8
NOTE: 数据集 WORK.TEST 有 1 个观测和 7 个变量。
NOTE: “DATA 语句”所用时间（总处理时间）:
      实际时间          0.03 秒
      CPU 时间          0.03 秒

For your second question, in sas there is a constant epsolon:

406  data _null_;
407  ep=constant('MACEPS');
408  put ep=;
409  run;

ep=2.220446E-16

Once the difference of two numbers is less than epsolon, then sas would take these two number are same, your situation is this.

For your third question, I also get surprise at this result. I also think

put(x1,5.1)

would be like this

put(round(x1,5.1),5.1)

and would expect both are 73.8 . Not one is 73.7 and another is 73.8 .

I suggest you to contact with SAS -Technical -Support to discover what is wrong with PUT() .

Re: 串聯檔案問題

Ksharp — Fri, 26 Jun 2026 06:58:02 GMT

这种情况下，UTF-8编码和wlatin1编码是不可以合并的。

因为有些字符在UTF-8中可以显示，在wlatin1中不可以显示。

你需要将这两个SAS数据集编码转为一样的，才可以合并。

Check SAS documentation macro:

https://go.documentation.sas.com/doc/en/pgmsascdc/9.4_3.5/nlsref/p1g1d26os4w0von1cdfh827foo3r.htm

https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.5/nlsref/p1f9ghftk3fgrin16t57ub6svmet.htm

Or my paper which is the attachment (PDF).

Re: Add white row before a variable value

Ksharp — Fri, 26 Jun 2026 06:46:08 GMT

Just have some fun.

data DB;
  input class :$20. subclass $20.;
cards;
DiseaseC1 Tot  
. variant1
. variant2
DiseaseC2  Tot  
. variant1
. variant12
. variant3
. variant15
;


data want;
 merge DB DB(keep=subclass rename=(subclass=_sub) firstobs=2);
 output;
 if _sub='Tot' then do;
  call missing(of _all_);output;
 end;
 drop _sub;
run;

Re: Setting the grid in SGPLOT but not setting an upper limit

Ksharp — Fri, 26 Jun 2026 08:09:14 GMT

Could you post dataset “druonly518” and the desired output graphic ? Without those ,it is hard to give you some advice.

If I understood what you mean, all you need to do is just adding one more option VALUESHINT

data have;
 set sashelp.heart;
run;


proc sort data=have(where=(bp_status is not missing)) out=have2;
by bp_status ; 
run;

proc sgplot data=have2; 
   by bp_status ;
   histogram weight /  GROUP=status scale=count transparency=0.5; 
   yaxis grid values=(0 to 1000 by 40) VALUESHINT;
run;

Re: odbc performance mystery--libname access slow, but sending same SQL via pass-through is really f

Ksharp — Thu, 25 Jun 2026 06:36:08 GMT

1)
If these two tables were not too big, I suggest you to copy them into sas side and do SQL's join statement.

proc copy in=vdw out=work noclone;
select utilization dx ;
run;

2)
Did you try option insertbuff= ?

Re: LSMeans i proc mix for subgroup category

Ksharp — Mon, 22 Jun 2026 06:16:22 GMT

You might need these ?

https://support.sas.com/kb/61/830.html

https://support.sas.com/kb/37/228.html

Re: Mean and Standard Deviation for Large Amounts of Data

Ksharp — Sat, 20 Jun 2026 14:44:52 GMT

I remembered @Rick_SAS wrote a blog about this topic. Check this:

https://blogs.sas.com/content/iml/2026/04/06/welford-mean-var.html

Re: Transposing Data

Ksharp — Thu, 18 Jun 2026 07:52:12 GMT

If your data include one customer has both NON_COE and COE rows, that would not be a problem, sas would generate two columns as you wanted.

But if your data include only NON_COE or COE row in each customer, that is another story, you need to pad both NON_COE and COE rows in it.

data have;
input customer flag $ count;
cards;
12345 NON_COE 2
12346 COE 5
;

data flag;
input flag $;
cards;
NON_COE
COE
;

proc sql;
create table have2 as
select a.*,coalesce(b.count,0) as count
 from 
(
select * from
(select distinct customer from have),
(select distinct flag from flag)
) as a natural left join have as b
;
quit;

proc transpose data=have2 out=want;
by customer;
var count;
id flag;
run;

Re: Autocorrelation in Repeated Measures Count Data with Overdispersion

Ksharp — Wed, 17 Jun 2026 07:10:53 GMT

1)Could you try Generalized Estimating Equations model ? Like:

PROC GENMOD

PROC GEE

2)It looks like you want to do panel data analysis. Check:

PROC PANEL

PROC COUNTREG

can be used for repeated count responses (panel data).

But they are all under SAS/ETS module , suggest you to post it at Forecasting Forum:

https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/bd-p/forecasting_econometrics

Re: Proc mixed constrast statement how to use specific examples

Ksharp — Tue, 16 Jun 2026 06:23:46 GMT

Here is an example how to write ESTIMATE for PROC MIXED if you want to compare random effect.

https://sas.service-now.com/csm/en/obtaining-subject-specific-parameter-estimates-and-tests-for-a-random?id=kb_article_view&sysparm_article=KB0057270

Re: SAS interruption and problems with reinstallation

Ksharp — Mon, 15 Jun 2026 07:17:54 GMT

Try to reinstall SAS software again ?

Re: Recurring question: Automatically Adjust Row Height in Excel

Ksharp — Thu, 11 Jun 2026 06:59:33 GMT

Chris,

Could you adjust it by hand ? Just click this button when you tap a cell.

Re: Recurring question: Automatically Adjust Row Height in Excel

Ksharp — Wed, 10 Jun 2026 07:06:08 GMT

Chris,

You could try option "tagattr=“ to avoid wrapping a line when you want to outpu an excel file.

Here is an example:

data HAVE;
  x = catx( '0D0A'x, '0000000000', '4445566778','9999999999','9991123456'); 
run;

ods excel file="c:\temp\print.xlsx" options(sheet_name='Test');          
  proc report data=HAVE nowd ;
  columns x;
  define _all_/display style={tagattr="wrap:no"};
  run;
ods excel close;

Re: Calculate time by comparing conditions and dates vertically

Ksharp — Tue, 09 Jun 2026 06:52:01 GMT

Assuming the data has been sorted by ID and DATE.

data DB;
  input ID :$200. Event:$200. Date :date09.;
    format Date date9.;
cards;
0001 CR    01SEP2024
0001 PR    10OCT2024
0001 CR    15DEC2024
0001 PD    22MAR2025
0001 CR    08AUG2025
0001 PD    20OCT2025
0001 Death 02DEC2025
0002 CR    01SEP2024
0002 PR    10OCT2024
0002 CR    15DEC2024
0002 PD    01MAR2025
0002 CR    08AUG2025
0002 PD    20OCT2025
;

data PD;
 set DB(where=(Event='PD'));
 by ID;
 if first.id;
run;

data other;
 set DB(where=(Event not in ('PD' 'Death')));
 by ID;
 if first.id;
run;

data want;
 merge PD other(rename=(event=event2 date=date2));
 by id;
 diff=date-date2;
run;

Re: Visitnumber derivation

Ksharp — Mon, 08 Jun 2026 07:25:26 GMT

Assuming I understood what you mean.

data have;
infile cards truncover expandtabs dlm=' ';
input USUBJID	(PARAMCD	ADTM) (:$40.)	AVISITN 	AVISIT	&$20. ADY	EOTDY		AVALC &$20. 	AVAL;
cards;
101	ABCD0102	2025-01-20T12:54:00	1	Week 1	 4	354	 	Slightly severe	 2
101	ABCD0102	2025-01-21T07:52:00	1	Week 1	 5	354	 	Slightly severe	 2
101	ABCD0102	2025-01-22T07:16:00	1	Week 1	 6	354	 	 	 
101	ABCD0102	2025-01-23T12:09:00	1	Week 1	 7	354	 	 	 
101	ABCD0102	2025-01-24T17:15:00	2	Week 2	 8	354	 	 	 
101	ABCD0102	2025-01-26T13:10:00	2	Week 2	 10	354	 	Severe	 3
101	ABCD0102	2025-01-28T11:09:00	2	Week 2	 12	354	 	Severe	 3
101	ABCD0102	2025-01-30T14:04:00	2	Week 2	 14	354	 	Slightly severe	 2
101	ABCD0102	2025-02-01T09:23:00	3	Week 3	 16	354	 	Slightly severe	 2
101	ABCD0102	2025-02-02T19:56:00	3	Week 3	 17	354	 	Slightly severe	 2
101	ABCD0102	2025-02-03T14:48:00	3	Week 3	 18	354	 	Slightly severe	 2
101	ABCD0102	2025-02-04T13:22:00	3	Week 3	 19	354	 	Slightly severe	 2
101	ABCD0102	2025-02-06T18:44:00	3	Week 3	 21	354	 	Slightly severe	 2
101	ABCD0102	2025-02-08T18:09:00	4	Week 4	 23	354	 	Slightly severe	 2
101	ABCD0102	2025-02-09T17:24:00	4	Week 4	 24	354	 	Not severe	 1
101	ABCD0102	2025-02-10T16:40:00	4	Week 4	 25	354	 	Not severe	 1
101	ABCD0102	2025-02-11T19:27:00	4	Week 4	 26	354	 	Not severe	 1
;

proc format;
invalue fmt
1=1
2=2
3=3
4=4
5=5
6=6
7=7
8=8
9=9
10=10
11=11
12=12
13-15=13
16-18=14
19-21=15
22-24=16
25-27=17
28-30=18
;
data have2;
set have;
id=input(AVISITN,fmt.);
run;
proc sql;
create table want as
select * from have2
 group by  USUBJID ,PARAMCD ,id
  having AVAL =max(AVAL );
quit;