Hello @SASdevAnneMarie,   Use blanks to separate the items in macro variable liste. The IN operator doesn't need commas and the macro processor doesn't like them.   Also omit the quotation marks around the value of nom_part because they are invalid characters in the name of a dataset or (macro) variable.   %my_analysis(liste='C1' 'C2',nom_part=PXXX) ... View more

@Yury2 wrote: Hello FreelanceReinh. Thank you for pointing this out. I would be more than happy to do this but when I click on the option menu on the left of the post I do not see the option "Not a Solution". I must be missing something. Please, advise. How strange. I've just accepted a solution to a thread I had started and this is how the option menu of the accepted solution looks like (red highlighting added): I'm not sure how long this option is available, but I think that at least today's solutions should have it. [Edit:] Indeed, this menu item is shown even for a solution I accepted more than eight years ago. ... View more

Hello @Yury2, Glad to see that antonbcristina's solution worked for you. Then it would be fair and help later readers if you marked her most helpful reply as the accepted solution, not your own "thank you" post. Could you please change that? It's very easy: Select her post as the solution after clicking "Not the Solution" in the option menu (see icon below) of the current solution. (The same goes for your other thread, by the way.) ... View more

Hello @alepage,   If you need this modified WEEKDAY function multiple times in your programs, you may want to create your own function ("MYWEEKDAY"): proc fcmp outlib=work.funcs.test; function myweekday(date); return(mod(weekday(date),7)+1); endsub; run;   Usage example: options cmplib=work.funcs; data _null_; set original_data; wd=myweekday(date); put wd date weekdate.-l; run; Result: 1 Saturday, December 28, 2013 2 Sunday, December 29, 2013 3 Monday, December 30, 2013 4 Tuesday, December 31, 2013 5 Wednesday, January 1, 2014 6 Thursday, January 2, 2014 7 Friday, January 3, 2014 ... View more

Hello @alepage,   It looks like the DATEKEYS procedure (in conjunction with the EVENTDS= system option) is designed for this purpose (i.e., adding a new holiday to the existing SAS-supplied holiday definitions). It is experimental in my fairly old SAS 9.4M5 release. /* Define a new holiday */ proc datekeys; datekeydef SAINT_JEAN_BAPTISTE_DAY='24JUN2000'd / period=year; /* The year (2000) is irrelevant. */ datekeydata out=JBdays; run; /* Make the new definition available */ options eventds=(JBdays); /* Use the new holiday like any other holiday */ data _null_; length holiday $50; do date='25DEC2020'd, '24JUN2021'd, '17APR2022'd; holiday=holidayname(date); put date :date9. holiday; end; d=holidaynx('SAINT_JEAN_BAPTISTE_DAY',today(),0); put / 'The most recent Saint-Jean-Baptiste Day was ' d worddate.-l +(-5) '.'; run; Result in the log: 25DEC2020 CHRISTMAS 24JUN2021 SAINT_JEAN_BAPTISTE_DAY 17APR2022 EASTER The most recent Saint-Jean-Baptiste Day was June 24, 2024. ... View more

Good luck to all contenders!   @ShelleySessoms wrote: Votes expressed as "likes" on entries here in the SAS Community will be factored in. However, mixing those votes and the likes given to normal contributions to the SAS Support Communities will result in a somewhat mixed Likes Leaderboard. The impact of last year's voting is still present in the 1-year statistics: Over the years, fourteen of the top 20 liked authors of the past 12 months have contributed a total of about 165,000 posts to the communities, i.e. more than 11,700 on average, whereas the remaining six have authored less than 3 posts on average. Most of the likes for their few posts are from community members whose activity doesn't seem to go much beyond submitting those votes. ... View more

@SteveDenham wrote: I would suggest the following code:   proc glimmix data=mydata plots=residualpanel method=laplace; /* you might want to explore method=quad or method=fastquad */ class group condition time; model Looking= group condition time group*condition group*condition*time /dist=beta link=logit ddfm=kr2 solution; random intercept / subject=ID type=un; random time / subject=ID(group*condition) type=un; run; (...) [Just to make the red highlighting visible, which was overridden by the automatic syntax highlighting.] ... View more

@SASJedi wrote: Using formats, we can see that 1117.11*100 is an irrational number in binary: (...) Variable Value Result y 111711 0100000011111011010001011110111111111111111111111111111111111111 Hi @SASJedi,   I like using the BINARY64. format in such situations. But I wouldn't call that number "irrational" because in mathematics this term has a special meaning (i.e., not being the ratio of two integers; cf. Wikipedia article Irrational number) which doesn't apply here. It's not even a repeating fraction (like 1/3=0.33333... in the decimal system), although the long string of 1s resembles 1/9=0.11111... in the decimal system. Since 64-bit floating-point numbers have only a finite number of binary digits (not more than 64 obviously) they are always rational.   The long string of 1s above is just the result of a rounding error: The decimal number 1117.11 written (mathematically) in the binary system is a (rational) repeating fraction: 10001011101.00011100001010001111010111000010100011110101110000101000111101... The 20-digit-pattern highlighted in blue is actually repeated forever (only three blocks of these are shown above). However, in the BINARY64. representation of that number (shown below) the infinite sequence is rounded to fit into the 64 bits, more precisely: into the 52 mantissa bits plus the implied bit (highlighted in green above). 0100000010010001011101000111000010100011110101110000101000111101 Due to the space restriction in the mantissa, the 20-digit-pattern is repeated only once, the next block "011100..." is rounded off.   Multiplying this number by the decimal number 100 (binary: 1100100) amounts to summing three shifted copies of it (because of the three 1s in "1100100"). Written mathematically as multiples of 2**10=1024: 1000101.1101000111000010100011110101110000101000111101 + 100010.11101000111000010100011110101110000101000111101 + 100.01011101000111000010100011110101110000101000111101 ----------------------------------------------------------- 1101101.00010111101111111111111111111111111111111111110101 Now we see how the round-off error mentioned above affects the summation and leads to a result that is slightly too small. Moreover, it contains 57 binary digits, so it must be rounded again to 52+1=53 bits: The "0101" at the end is rounded down (because of the leading zero), leaving that long (but not infinite) string of 1s in the mantissa.   @cklager44: "If the argument is within 1E-12 of an integer, the INT function fuzzes the result to be equal to that integer." (INT function documentation) Unfortunately, the difference between the result and the exact integer 111711 exceeds 1E-12, as Tom has shown already (diff=1.4551...E-11).   The situation with 1117.01*100 is actually similar, but in the last rounding step -- from 57 to 53 bits -- we are just lucky: Instead of "0101" (as above from 1117.11*100) a "1111" needs to be rounded and this is rounded up (because of the leading 1), which compensates the previous rounding error and leads to the proper internal binary representation of the integer 111701. ... View more

@quickbluefish wrote: It's possible the correct formula for Ecc just involves squaring the inner part (thus guaranteeing a positive number) before taking the square root, in other words, the just taking the absolute value, though not sure it wouldn't just be written with absolute value notation if that were the case.  This would be "possible" (albeit unusual) if we knew nothing about the context of the formula. But the corresponding formula in the Wikipedia article and its explanation clearly show that an opening parenthesis must be inserted after the minus sign: Ecc=sqrt(1-(WtHR/3.14159)**2); Only this correction fits the definition of the eccentricity of an ellipse in geometry: see the first table in Wikipedia article Eccentricity (mathematics). ... View more

Hello @GaryHeathcote and welcome to the SAS Support Communities!   @GaryHeathcote wrote: WtHR=round((waistcir/stature)*100, .1);   Ecc=sqrt(1-WtHR/3.14159)**2);   BRI=(364.2 - 365.5)*Ecc; I agree with the other experts that you should first make sure your formulas are correct before implementing them in code.   It is possible that all three of the above formulas are incorrect: Comparing them to the similar three-step approach presented in the Wikipedia article Body roundness index, your factor 100 in the first formula would make sense if waistcir was measured in meters and stature in centimeters. But is this really the case (and not vice versa)? If not, this would most likely cause the "Invalid argument to sqrt" problem (because then WtHR [which Wikipedia denotes WHtR] would tend to be way too large). The rounding unit .1 should be checked, too: Rounding WHtR to tenths in the numeric example shown in the Wikipedia article would change the result quite significantly (from 2.47 to 3.36), which is probably not desirable. In the second formula the opening parenthesis after the minus sign is missing (as others have mentioned already). The parentheses in the third formula are definitely wrong and must be removed. (As posted, the BRI would always be negative and the formula could obviously be simplified.) ... View more

Bonjour @SASdevAnneMarie,   data test; input lb_c $20.; crit1=prxmatch("/T6/",lb_c) and prxmatch("/T 6/",lb_c); crit2=prxmatch("/T6/",lb_c) or prxmatch("/T 6/",lb_c); label crit1='Contains both "T6" and "T 6"' crit2='Contains "T6" or "T 6" (or both)'; cards; ...... ...T6... ...T 6... ...T 6...T6... ...T6...T 6... ; proc print data=test label; run; Result: Contains Contains "T6" or both "T6" "T 6" Obs lb_c and "T 6" (or both) 1 ...... 0 0 2 ...T6... 0 1 3 ...T 6... 0 1 4 ...T 6...T6... 1 1 5 ...T6...T 6... 1 1 ... View more

Here's another just-for-fun approach, suitable (without further additions) only for small examples like your sample data or sashelp.class, if any: proc sql noprint; select catx(':',department,employee,department) into :x separated by '|' from employees order by department, employee; create table want as select department, scan(substr(tranwrd("&x",cats(':',department,'|',department,':'),', '),find("&x",cats(department,':'))),2,':') as emp_list from (select distinct department from employees); quit; ... View more

Your encouraging comments, dear Reeza, helped me a lot when I started posting here nine years ago. Thank you so much for your support and for what I learned from you.   Your tremendous contributions to the SAS Support Communities are reflected by your top positions in both the Solution Authors Leaderboard and the Likes Leaderboard (where another community giant took over the all-time No. 1 spot only very recently). They will doubtlessly continue to help many SAS users around the globe.   It's reassuring to know that there is a professional life after SAS. May success follow you in your post-SAS career.   Thanks again and best wishes,   Reinhard ... View more

The community manager of the German forum has just posted that their service provider has found the bug and will fix it soon. ... View more