You're welcome.   Those macros are closely linked to the name of SAS employee Warren F. Kuhfeld, PhD (now retired), see his initials in the header of the code (where also the "update" date July 25, 2013 documents how long they have been available and certainly used by many SAS programmers around the globe). So I assume that the macros were developed by himself or under his guidance. At least, he is the author of the documentation, as you can read in https://blogs.sas.com/content/author/warrenkuhfeld/ together with more information about him. He also wrote the 2017 blog article Customizing the Kaplan-Meier Plot and answered related questions in the SAS Support Communities, e.g., Re: Format KM curve with median survival line.   I can't imagine that SAS Institute would provide such macros in their official documentation if they weren't validated. ... View more

Thanks, @yabwon, for testing.   It seems to me that restrictions are a big topic when it comes to CAS (or FedSQL for that matter) ...   PROC SQL Statement Restriction: This procedure is not supported on the CAS server. ALTER TABLE Statement (...) Restrictions: This statement is not supported in FedSQL programs that run in CAS. ALTER [COLUMN] <column-definition> [, ...<column-definition>] specifies to modify the definition of one or more columns. Data source Amazon Redshift, Aster, DB2 under UNIX and PC, ... [and several others, but NOT including "SAS data set"] ... View more

I don't have CAS available here. Just wondering, wouldn't this work in CAS? proc sql; alter table myCas.cars(replace=YES) modify model char(128); quit; ... View more

Hello @charleenjm,   @charleenjm wrote: ... I run into an error "Array subscript out of range at line 172 column 12. ..." This error message occurred because for at least one observation in dataset SCDNY2016 the IF condition in the loop checking for sickle cell disease was never met and hence the LEAVE statement was not executed. As a consequence, the DO loop was left with the end value plus 1  in variable i, that is i=dim(dx_vars)+1=26. Indeed, the log contains "i=26" (but it would be easier to find if you posted the log as text, not as a screenshot).   It appears that you forgot the DO-END statement in (what is announced to be) the "loop ... to check for delivery diagnoses", so that no loop is executed here and only the leftover value i from the previous DO loop is used. If that value is too large (as explained above), you get the error message when referring to dx_vars[i] in the next IF statement. Otherwise, the log might be clean, but the result is likely wrong.   The missing underscore, well spotted by @Kathryn_SAS, must have caused another error message (but you didn't post the entire log of the DATA step, so we don't see it), while the dollar signs in the array definitions are optional for existing character variables. ... View more

Yes, this implication is always true for two binary random variables A and B on some probability space. Even either of the two equations @TomHsiung wrote: P (A = 1 & B = 0) = P(A=1)*P(B=0) P (A = 0 & B = 1) = P(A=0)*P(B=1) alone implies the other and the remaining two combinations P (A = 1 & B = 1) = P(A=1)*P(B=1) and P(A=0 & B=0) = P(A=0)*P(B=0), all of which mean that A and B are independent.   For example, assuming P(A=1 & B=0) = P(A=1)*P(B=0), we obtain P(A=1 & B=1) = P(A=1) - P(A=1 & B=0) = P(A=1) - P(A=1)*P(B=0) = P(A=1)*(1-P(B=0)) = P(A=1)*P(B=1). ... View more

Hi @NewUsrStat,   You can also use the MATCH_ALL option available in the ODS OUTPUT statement: /* Create sample data for demonstration */ data have; call streaminit(27182818); array a[5] Q17_test Q17_new Q17_tmp Q17_abc Q17_314; do i=1 to 51; do _n_=1 to dim(a); a[_n_]=rand('integer',9); end; output; end; run; /* Create datasets _TMP1, _TMP2, ... containing a frequency table for each "Q17_" variable */ ods output onewayfreqs(match_all)=_tmp1(drop=Table F_: Cum: Percent rename=(Frequency=COUNT)); proc freq data=have; tables Q17_:; run; /* Compile pairs of dataset names and prefixed variable names */ proc sql noprint; select cats('_tmp',varnum,'=stats_',name) into :chg_list separated by ' ' from dictionary.columns where libname='WORK' & memname='HAVE' & upcase(name) eqt 'Q17_'; quit; /* Rename the ODS output datasets to STATS_<variable name> */ proc datasets lib=work nolist; change &chg_list; quit; ... View more

Are there cases in your data where num_PW=input(PW, 32.) does not contain the "correct decimal number", that is, where a check like 371 data _null_; 372 PW="7.0000000000000007E-2"; 373 num_PW=input(PW, 32.); 374 if num_PW=0.07 then put 'OK'; 375 run; OK would not yield "OK"? If so, I would use the ROUND function with an appropriate rounding unit in the definition of num_PW.   Example: 412 data _null_; 413 PW="7.0000000000000010E-2"; 414 num_PW=round(input(PW, 32.), 1e-14); 415 if num_PW=0.07 then put 'OK'; 416 run; OK   ... View more

Hello @abigail_greenh,   To convert the 21-character string "7.0000000000000007E-2" to a numeric value, use an informat with a length >=21, for example the 32. informat:   num_PW=input(PW, 32.);   The informat BEST12. (which is just an alias of 12.) has only length 12 and hence would miss the important "E-2" information, resulting in the integer value 7. Also note that the "w " (written in italics) in format and informat names found in the SAS documentation must always be replaced by a numeric width specification (and "d " by a number of decimals, but these are not needed in your example).   I can confirm that 0.07 is stored as 7.0000000000000007E-2 in an .xlsx file created with Excel 2013 on Windows 10. After unzipping the underlying xml files of an Excel sheet containing the numbers 0.01, 0.02, ..., 0.11 I see that, among these 11 numbers, only 0.07 has that precision issue. Here is an excerpt of sheet1.xml:   ... <row r="6" spans="1:1" x14ac:dyDescent="0.25"><c r="A6"><v>0.06</v></c></row> <row r="7" spans="1:1" x14ac:dyDescent="0.25"><c r="A7"><v>7.0000000000000007E-2</v></c></row> <row r="8" spans="1:1" x14ac:dyDescent="0.25"><c r="A8"><v>0.08</v></c></row> ...   I found several discussions of this Excel problem on the web from 2009 and later years.   Addendum: By the way, the internal binary representation of 0.07 in SAS on Windows and Unix platforms, mathematically translated back to the decimal system, equals 0.070000000000000006661338147750939242541790008544921875 Rounded to 18 decimals, this matches the value 7.0000000000000007E-2 seen in Excel.   Addendum 2: SAS (at least my Windows SAS 9.4M5) would interpret the numeric literal 7.0000000000000007E-2 as the same value as 0.07, 45 data _null_; 46 if 7.0000000000000007E-2=0.07 then put 'OK'; 47 run; OK although the internal representation corresponding to that decimal fraction, derived mathematically, would have a numeric value slightly greater than 0.07 (see this 2024 post for more details). ... View more

@Aashi07 wrote: I am using Unix This is not a problem as all functions used in my suggested code are available in Unix SAS, too. Of course, the path will contain forward slashes instead of backslashes and the "last modified" datetime format may be different. In a preliminary step, you could read that datetime into a character variable  like this lastmod=finfo(fid,'Last Modified'); and then look at the character values in variable LASTMOD to find out the appropriate date or datetime informat to be used eventually.   See also Example 4 of the FINFO function documentation where the last modification datetime, among other file information items, is retrieved on a Unix system.   If you encounter difficulties, please provide details such as the complete log of the DATA step that you tried. ... View more

Hello @Aashi07,   This works on my Windows workstation: /* Create sample data for demonstration */ data have; input name :$200.; cards; test.sas7bdat test0.lst test1.log test2.sas ; /* Check existence and retrieve last modification date of the above files in a specified folder */ %let folder=C:\Temp\test; data want; length path $260 fref $8; set have; path="&folder\"||name; exists=fileexist(path); if exists then do; rc=filename(fref,path); fid=fopen(fref); lastmod=input(finfo(fid,'Last Modified'),nldate200.); rc=fclose(fid); end; format lastmod yymmdd10.; keep path exists lastmod; run; Increase the lengths of variables NAME and PATH if necessary. Note that the "last modified" date is actually a datetime, so you could read it with the nldatm200. informat (in the INPUT function call) and use a datetime format such as e8601dt. in the FORMAT statement. ... View more

@TomHsiung wrote: I think the last two formula means A = 1 and B = 0 as well as A = 0 and B = 1 are independent, respectively. That's why I'm confused about the independence between event A and B. Hello @TomHsiung,   I see your point.   My understanding is that the random variables describing the individual responses are independent, hence the products of probabilities in the Wikipedia article you have mentioned. Yet, as a rule, matched pairs are correlated in the following sense: If X, Y denote the responses of a randomly selected matched pair, these random variables X and Y are usually dependent because they tend to have similar response probabilities due to the matching.   Here is a small example with only two matched pairs: (X 11 , X 12 ) and (X 21 , X 22 ), each representing, say, (case, control). Assume independent Bernoulli distributions X 11 ~B(1, 0.94), X 12 ~B(1, 0.91), X 21 ~B(1, 0.22), X 22 ~B(1, 0.29).   Define X:=X U1 , Y:=X U2 with a random variable U, independent of the X ij , describing the selection of a pair: P(U=1)=P(U=2)=0.5.   Then we obtain the joint distribution of (X, Y) from calculations like P(X=1, Y=1) = P(U=1)P(X 11 =1)P(X 12 =1)+P(U=2)P(X 21 =1)P(X 22 =1)=0.4596:   P X=0 X=1 Y=0 0.2796 0.1204 0.4 Y=1 0.1404 0.4596 0.6   0.42 0.58 1   Now we see that X and Y are correlated, hence dependent: Their correlation coefficient is r X,Y =(0.4596-0.58*0.6)/sqrt(0.58*0.42*0.6*0.4)=0.46155... ... View more

There doesn't seem to be an option of the PANELBY statement to control that space. In this case you could specify the NOHEADER option and create your own header using SG annotation. Not sure if this would be worth the effort, though. ... View more