"  %include  doesn’t work well inside a macro like that because it runs before the macro finishes. " - please explain what do you mean by this sentence? Because in the form it is, it is false.   Bart ... View more

or compbl() function ... View more

Multiple spaces can be a bit of a problem here:   %let kept=Arne Bert Hansi; gives 46 data _null_; 47 call symputx("w_commas",tranwrd(quote(tranwrd("&kept",' ',quote(', '))),'""','"')); 48 call symputx("wo_commas",tranwrd(quote(tranwrd("&kept",' ',quote(' '))),'""','"')); 49 run; NOTE: DATA statement used (Total process time): real time 0.01 seconds cpu time 0.01 seconds 50 51 %put &=w_commas; W_COMMAS="Arne", "", "", "", "Bert", "", "Hansi" 52 %put &=wo_commas; WO_COMMAS="Arne" "" "" "" "Bert" "" "Hansi"   Bart ... View more

One more option just for fun, with FCMP function called in %SYSFUNC() proc fcmp outlib=work.f.p; function addQuotes(str $,s $, q) $ 32767; length result $ 32767 sep word Usep $ 128 qt $ 1; sep = s; Usep = upcase(sep); select; when (Usep=" ") sep=" "; when (Usep="C") sep=","; when (Usep="SC") sep=";"; when (Usep="P") sep="."; when (Usep=:"S") sep=substr(sep,2); /* string up to 127 bytes */ otherwise sep = " "; end; select(q); when(2) qt='"'; when(1) qt="'"; otherwise qt=" "; end; quit=0; do i = 1 by 1 while(NOT quit); word = scan(str, i, " "); if NOT missing(word) then do; if NOT (qt=" ") then word=quote(strip(word),qt); result = catX(ifc(sep=" "," ",strip(sep)), result, strip(word)); end; else quit=1; end; return(result); endfunc; quit; options cmplib=work.f; %let kept=Arne Bert Hansi; %put %sysfunc(addQuotes(&kept.,C,2)); %put %sysfunc(addQuotes(&kept.,S!,0)); data _null_; kept="Arne Bert Hansi"; str=addQuotes(kept,"C",2); put kept= / str=; str=addQuotes(kept,"SC",0); put kept= / str=; run;   Bart ... View more

O course DoSubL() function can help in wrapping it into a macro: %let kept=Arne Bert Hansi; %macro addQuotes(kept); %local rc quoted_names; %let rc=%sysfunc(dosubl(%str( data _null_; length str result $ 32767 word $ 128; str=symget("kept"); sep = " "; /* space */ do i=1 by 1 while(NOT quit); word = scan(str, i, " "); if word NE " " then result = catX(sep, result, quote(strip(word)) ); else quit=1; end; call symputX("quoted_names", result, "L"); run; ))); &quoted_names. %mend addQuotes; %put %addQuotes(&kept.); Bart ... View more

do you need macro for the process? how about data step?   %let kept=Arne Bert Hansi; data _null_; length str result $ 32767 word $ 128; str=symget('kept'); sep = " "; /* space, but you can set whatever you like */ do i=1 by 1 while(NOT quit); word = scan(str, i, " "); if word NE " " then result = catX(sep, result, quote(strip(word)) ); else quit=1; end; call symputX("quoted_names", result, "G"); run; %put &quoted_names.;   Bart ... View more

If someone doesn't have IML, just for fun, 4GL version (there is a bunch of code commented out from my try-and-errors for optimisation, for those who like to test):     /* arrays simulation */ options mprint; data _null_; array bwb[24] _temporary_ (8*(1 2 3)); /* template box with balls */ array b[24] _temporary_; /* box from which we are selecting */ array r[3] ; /* couter for selected color 1, 2, 3 */ call streaminit(42); /* set seed = 42 */ array x[0:12,0:12,0:12] _temporary_; /* array for results, 0=color was never selected */ do iter=0 to 1e6-1; /* iterations */ /* populate b with balls */ /* do n=1 to 24; b[n]=bwb[n]; end; *//* call pokelong ( put (PEEKCLONG (ADDRLONG(bwb[1]), 192), $192.) , ADDRLONG(b[1]) , 192); *//**/ %macro loop(n); %do n=1 %to &n.; b[&n.]=bwb[&n.]; %end; %mend loop; %loop(24) /**/ r1=0;r2=0;r3=0; /* initiate counters */ /* do j=24 to 13 by -1; i = RAND("integer",1,j); r[b[i]]+1; b[i]=b[j]; end; */ %macro loop2(n,m); %do n=&n. %to %eval(&n.-&m.+1) %by -1; /* select 12 from 24 with no repetitions */ i = RAND("integer",1,&n.); /* get random number from 1-24, 1-23, 1-22, etc. */ r[b[i]]+1; /* increase count of selected ball */ b[i]=b[&n.]; /* fill gap in "i" (after selected ball) with last ball ("n") */ %end; %mend loop2; %loop2(24,12) call sortn(of r[*]); /* order counters */ x[r1,r2,r3]+1; /* increase selected "set" */ end; /* output generated data */ do r1=0 to 12; do r2=0 to 12; do r3=0 to 12; if x[r1,r2,r3] then /* print only "materialized" cases */ do; n=x[r1,r2,r3]/iter; output; put @2 r1= @12 r2= @22 r3= @32 n= best32.20; end; end; end; end; keep r: n; run; Result for seed=42 and 1e9 iterations:   /* 1e9 r1=0 r2=4 r3=8 n=0.000155304 r1=0 r2=5 r3=7 n=0.000993489 r1=0 r2=6 r3=6 n=0.000869773 r1=1 r2=3 r3=8 n=0.000994072 r1=1 r2=4 r3=7 n=0.009935598 r1=1 r2=5 r3=6 n=0.027825111 r1=2 r2=2 r3=8 n=0.000869628 r1=2 r2=3 r3=7 n=0.027822129 r1=2 r2=4 r3=6 n=0.121775441 r1=2 r2=5 r3=5 n=0.097433405 r1=3 r2=3 r3=6 n=0.097416671 r1=3 r2=4 r3=5 n=0.487068081 r1=4 r2=4 r3=4 n=0.126841298 NOTE: The data set WORK.FREQCOUNT has 13 observations and 4 variables. NOTE: DATA statement used (Total process time): real time 11:56.18 cpu time 11:56.14 */   Bart ... View more

@FreelanceReinh  Looks like we went the same path 😄 ... View more

Just to be clear: probability is not my super-power but:   Ain't that case of multivariate hypergeometric distribution? https://en.wikipedia.org/wiki/Hypergeometric_distribution#Multivariate_hypergeometric_distribution and the fact that: P(3,4,5) is independent from P(3,5,4), etc, P(3,4,5 or 3,5,4) = P(3,4,5) + P(3,5,4) - P(3,4,5 and 3,5,4) = P(3,4,5) + P(3,5,4) so probability P("I have 3:4:5 sequence") would be just sum of P(3,4,5) + P(3,5,4) + ...+ P(5,4,3).     data have; input grp $ r g b; cards; a 4 4 4 b 3 4 5 b 3 5 4 b 4 3 5 b 4 5 3 b 5 3 4 b 5 4 3 ; run; /* a = get all 4 b = get 3:4:5 */ data want; rn=8; gn=8; bn=8; N=rn+gn+bn; set have; k=r+g+b; p = /* https://en.wikipedia.org/wiki/Hypergeometric_distribution#Multivariate_hypergeometric_distribution */ (comb(rn, r) *comb(gn, g) *comb(bn, b) ) /comb(n, k) ; put (_ALL_) (=/); if _N_=2 then sum=0; sum+p; run; proc print; run;   [EDIT:] proc print:       Bart   ... View more

Something like this will produce result for all matches: single, double or even triple: data data1; input ID VAR1 $; datalines; 1 A 1 B 1 C 2 X 2 Y 3 P 3 Q ; run; data data2; input ID1 ID2 ID3 VAR2; datalines; 1 . . 22 1 . . 25 4 2 . 40 4 2 . 24 4 2 . 29 5 6 3 12 5 6 3 15 1 2 . 40 1 2 3 42 ; run; data want; if 1=_N_ then do; if 0 then set data1; declare hash H(dataset:"data1", multidata:"Y"); H.defineKey("id"); H.defineData("var1"); H.defineDone(); drop id; end; array xx id1-id3; set data2; do over xx; if xx then do while(H.do_over(key:xx)=0); output; end; end; run; proc print; run;     Bart ... View more

What result would you expect in case when "data2" has: 1 2 . 40 1 2 3 42 in its observations?    Bart ... View more

@FreelanceReinh , 100% agree with your opinion, the SASDocViewer is 1e6-times "nicer" to work with then the doc page. I had situations when  while clicking through SASDocViewer I was able faster access help info than loading a googled page. The biggest advantage of the SASDocViewer is that it's related directly to SAS version you are working on. When one uses EG or Studio, and searches doc website, sometimes a "versions mix" situation happens, when work is on SAS version "X" and the doc page is for "Y".  In 2019 SFG in Dallas, during live testing of EG8 and SAS studio, I've suggested SAS should add "local" help to those tools...   Bart ... View more

Maybe it's like the last message in this thread:  https://communities.sas.com/t5/All-Things-Community/Bulkload-with-Redshift/m-p/470173#M3253 ?   Bart ... View more

It looks like your AWS connection setup need credentials with key & secret.   According to this doc. note: https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.5/acreldb/p04eb8ihn5xe8in1lzdxep6opttr.htm "Amazon Redshift, Snowflake: The secret-access-key value specifies the Amazon Web Services (AWS) secret access key or a temporary secret access key. An AWS secret access key is associated with the key ID that you specified with the BL_KEY= data set option. If you are using temporary token credentials, this value is the temporary secret access key."    Maybe ask your admins team about details for credentials?    Bart ... View more