About NKormanik

NKormanik · ‎08-02-2013

Thanks Vincent , , for the input. Vincent asks, "How do we measure performance in each round?" Assume only what's given -- the 'Order' the 'Pair' came in for each 'Round'. Seems reasonable to sum 'Order' across all 'Rounds' for each individual, and compare grand sums, ascending, for overall rank orderings. It should be pointed out that the data set here was created just for this post. The actual data set is far larger -- more than 100 'Contestants' x 100 'Rounds'. Thus Astounding's suggestion may not be practicable, because the resulting numbers of combinations would be so large. I'm not sure how to code it anyway. If Astounding or Steve wants to take a crack at it..... Super. A "race against the clock" is one good way of viewing the contest. Thanks Vincent. For pairs..., lowest time sum wins. Pairs, along this line, could be broken down as well. Lowest grand time sum wins. Of course, we lack the 'Time' data. Only have the 'Rounds' rankings. https://communities.sas.com/people/Vince28%40Statcan

NKormanik · ‎08-01-2013

Last time I asked the question with SINGLE individuals competing in each of several events, and how to determine the overall ranking of winners. https://communities.sas.com/thread/44071 (thanks , , , , ) Now I return with things a little more complicated: Suppose we have PAIRS in each contestant group. I would like to find a way of ranking the top INDIVIDUALS, even though they are taking part as a PAIR. In this competition, every individual was paired up round-robin, with every other individual. This was repeated seven times.... over a period of one year. We have the list of winners for each round, seven rounds in all. Unfortunately some pairs missed certain rounds. Thus, not a complete and even data set. One might guess that the top PAIR would include the, say, #1 and #2 contestants overall. My hunch is that such need not be the case. Could be that a single contestant from each of Pair 1, Pair 2, Pair 3 might be the top rankers overall. So, again, what we want is a ranking list of all INDIVIDUALS, based on the data set. Any help and advice is greatly appreciated. Thanks, Nicholas Kormanik ======= Edit by Nicholas: I've added a new data file. In it I've separated the 'Pairs' into two columns each, so that one needn't do it in SAS. 'Round 1' becomes 'R1a' and 'R1b', etc. I tried using Vince28's code and kept getting SAS errors. Bummer. Maybe by separating the 'Pairs' the task can be handled more easily. Again, any help greatly appreciated. ======= Data set follows (and additionally as an attachment): Order Round 1 Round 2 Round 3 Round 4 Round 5 Round 6 Round 7 1 DREW_EDGAR HENRY_ISAAC EDGAR_GRANT CHUCK_GRANT HENRY_ISAAC EDGAR_ISAAC AARON_FRANK 2 GRANT_ISAAC CHUCK_HENRY BLAKE_DREW GRANT_HENRY AARON_EDGAR EDGAR_HENRY DREW_EDGAR 3 DREW_HENRY DREW_ISAAC AARON_BLAKE BLAKE_CHUCK FRANK_HENRY AARON_ISAAC CHUCK_DREW 4 CHUCK_HENRY AARON_EDGAR AARON_CHUCK DREW_GRANT AARON_DREW BLAKE_CHUCK FRANK_GRANT 5 CHUCK_FRANK FRANK_ISAAC AARON_GRANT AARON_GRANT DREW_GRANT FRANK_GRANT AARON_EDGAR 6 AARON_EDGAR DREW_FRANK BLAKE_ISAAC EDGAR_GRANT BLAKE_FRANK DREW_FRANK EDGAR_ISAAC 7 BLAKE_GRANT CHUCK_FRANK DREW_GRANT DREW_HENRY BLAKE_HENRY BLAKE_ISAAC FRANK_HENRY 8 AARON_CHUCK AARON_ISAAC GRANT_ISAAC EDGAR_ISAAC BLAKE_GRANT DREW_ISAAC BLAKE_HENRY 9 EDGAR_GRANT AARON_HENRY BLAKE_CHUCK AARON_BLAKE AARON_ISAAC FRANK_HENRY AARON_BLAKE 10 CHUCK_ISAAC FRANK_HENRY AARON_FRANK BLAKE_ISAAC CHUCK_ISAAC AARON_CHUCK CHUCK_ISAAC 11 AARON_BLAKE DREW_GRANT BLAKE_HENRY EDGAR_HENRY CHUCK_FRANK DREW_HENRY CHUCK_HENRY 12 BLAKE_CHUCK BLAKE_ISAAC AARON_EDGAR DREW_FRANK DREW_ISAAC AARON_GRANT EDGAR_GRANT 13 FRANK_HENRY BLAKE_EDGAR FRANK_ISAAC EDGAR_FRANK FRANK_ISAAC GRANT_HENRY BLAKE_DREW 14 BLAKE_DREW EDGAR_GRANT DREW_HENRY DREW_EDGAR DREW_HENRY EDGAR_GRANT CHUCK_EDGAR 15 BLAKE_HENRY AARON_DREW DREW_ISAAC AARON_FRANK AARON_HENRY BLAKE_EDGAR AARON_ISAAC 16 CHUCK_DREW CHUCK_GRANT BLAKE_GRANT CHUCK_EDGAR BLAKE_CHUCK BLAKE_DREW DREW_GRANT 17 FRANK_GRANT CHUCK_ISAAC GRANT_HENRY CHUCK_DREW EDGAR_ISAAC DREW_EDGAR BLAKE_CHUCK 18 FRANK_ISAAC AARON_CHUCK EDGAR_FRANK CHUCK_HENRY EDGAR_GRANT AARON_FRANK GRANT_HENRY 19 GRANT_HENRY BLAKE_HENRY AARON_ISAAC FRANK_GRANT CHUCK_GRANT BLAKE_HENRY BLAKE_ISAAC 20 DREW_FRANK AARON_GRANT FRANK_HENRY BLAKE_DREW DREW_FRANK CHUCK_GRANT DREW_ISAAC 21 BLAKE_ISAAC BLAKE_DREW AARON_HENRY CHUCK_FRANK CHUCK_DREW CHUCK_ISAAC AARON_CHUCK 22 AARON_DREW GRANT_ISAAC CHUCK_EDGAR HENRY_ISAAC GRANT_HENRY AARON_EDGAR GRANT_ISAAC 23 AARON_GRANT DREW_EDGAR CHUCK_GRANT GRANT_ISAAC CHUCK_HENRY CHUCK_DREW EDGAR_HENRY 24 HENRY_ISAAC EDGAR_HENRY CHUCK_FRANK CHUCK_ISAAC CHUCK_EDGAR HENRY_ISAAC BLAKE_EDGAR 25 EDGAR_HENRY FRANK_GRANT DREW_FRANK BLAKE_GRANT AARON_FRANK AARON_HENRY HENRY_ISAAC 26 CHUCK_EDGAR AARON_FRANK BLAKE_FRANK AARON_HENRY GRANT_ISAAC CHUCK_FRANK BLAKE_GRANT 27 EDGAR_ISAAC DREW_HENRY EDGAR_HENRY FRANK_ISAAC AARON_GRANT FRANK_ISAAC BLAKE_FRANK 28 CHUCK_GRANT AARON_BLAKE CHUCK_HENRY AARON_DREW FRANK_GRANT CHUCK_HENRY FRANK_ISAAC 29 DREW_GRANT CHUCK_EDGAR HENRY_ISAAC BLAKE_FRANK BLAKE_EDGAR AARON_DREW AARON_DREW 30 DREW_ISAAC CHUCK_DREW CHUCK_ISAAC AARON_CHUCK BLAKE_ISAAC DREW_GRANT DREW_FRANK 31 AARON_HENRY GRANT_HENRY CHUCK_DREW DREW_ISAAC DREW_EDGAR BLAKE_FRANK CHUCK_GRANT 32 BLAKE_FRANK EDGAR_ISAAC BLAKE_EDGAR AARON_ISAAC AARON_BLAKE CHUCK_EDGAR AARON_HENRY 33 AARON_FRANK BLAKE_FRANK DREW_EDGAR BLAKE_HENRY EDGAR_HENRY BLAKE_GRANT DREW_HENRY 34 BLAKE_EDGAR EDGAR_FRANK EDGAR_ISAAC FRANK_HENRY AARON_CHUCK AARON_BLAKE AARON_GRANT 35 AARON_ISAAC BLAKE_GRANT AARON_DREW * BLAKE_DREW GRANT_ISAAC CHUCK_FRANK 36 * BLAKE_CHUCK FRANK_GRANT * EDGAR_FRANK * EDGAR_FRANK

NKormanik · ‎05-15-2013

, I appreciate you were trying to get me to learn. Thanks. It would be great, though, just for completeness for anyone else coming across this thread, if you could put the rest of the code. The first part seemed beautifully written.

NKormanik · ‎05-15-2013

, I mean in the cards section at top..., can I include 50 lines?

NKormanik · ‎05-15-2013

, by the way, might as well ask now..., would it be okay to include the top 50 rows using your code? Or is it just limited to 20, as requested in the original post? I'll try. But just wanted to ask you now since it's before you presently. Thanks!

NKormanik · ‎05-15-2013

Beautiful, beautiful, beautiful!! Thanks soooo much. Nicholas

NKormanik · ‎05-15-2013

Oops. I made a mistake there. Was trying to fill it in manually. Not an easy transformation. Well..., that's exactly the reason for wanting the SAS code to do this, so that the job gets done properly. I've gone back and tried to make corrections. Thanks, , for noticing and pointing out the error.

NKormanik · ‎05-14-2013

I've tried repeatedly to follow what you provided in the first go-round. But with my actual data, it just didn't work. . So, I'll give the actual data here, and beg for more help. Suppose we are looking at 'marathon' results. Assume 5 'races' by hundreds of 'runners'. We only examine the top 20 finalists (but perhaps the top 30, or top 50). Here is the order they come in for the 5 races.... . Order Race_1 Race_2 Race_3 Race_4 Race_5 1 22604_22804 23401_23403 23401_23403 23401_23403 23401_23403 2 22402_22604 20506_23003 22604a_23005 22702_23003 20506_21402 3 22601_22604a 22604a_23005 20506_23003 20506_21402 21802_23401 4 20206_21503 21402_23401 20506_21402 21702_23306 21402_21805 5 21503a_22002 21802_23401 21702_23306 21402_22706 20801_23003 6 22402_23003 21702_23306 22702_23003 21801_22603a 22805_23002 7 20706_21605 21402_23403 21005_21301 20505_21402 21402_21806 8 20901_22803 20801_22604a 20801_23004 21402_21805 22702_23003 9 21402_22804 20506_21402 20801_22604a 21402_22705 21402_22706 10 21401_23101 23002_23403 20505_21402 20506_23003 21702_23306 11 20301_20501 21806_23004 21805_23003 20801_23004 21802_23405 12 20305_20901 20801_23004 21801_22603a 22604a_23005 21801_22603a 13 20301_23005 22301_23003 22604a_23002 22604a_23002 22001_23003 14 21502_21806 21005_21301 21402_22706 20301_21801 20505_21402 15 20802_21005 21103_21801 22703_23003 21402_21806 21402_21703 16 21602_22804 22703_23003 21402_23401 21402_22703 21802_23403 17 21004_21401 23002_23401 20301_21801 22703_23003 21402_22703 18 20203_22401 20505_21402 21402_22705 22706_23003 21402_22705 19 21605_23202 21402_22705 21402_23403 21401_21501 20801_21402 20 21502_22602a 21402_23405 21806_23003 21805_23003 21402_23401 . I need to transform the above data format to show the Runner in the first column, followed by the Order each runner comes in for the particular race.... . Runner Order_1 Order_2 Order_3 Order_4 Order_5 20506_21402 * * * 3 2 20506_23003 * 2 3 10 * 22402_22604 2 * * * * 22601_22604a 3 * * * * 22604_22804 1 * * * * 22604a_23005 * 3 2 12 * 22702_23003 * * 6 2 8 23401_23403 * 1 1 1 1 . Note that 23401_23403 placed first in the last four races, but did not even place in the top 20 in the first race. . The reason for doing the above transformation is so that I can run Proc Prinqual -- roughly the same example as: . Example 74.2 Principal Components of Basketball Rankings SAS/STAT(R) 12.1 User's Guide . I'll greatly appreciate your help. . Thanks. . Nicholas Kormanik .

NKormanik · ‎05-03-2013

Here is the contribution provided by given on another thread. This translates the Basketball example which uses PCA earlier mentioned by : data input; input rank (race1-race3) ($) ; cards; 1 D E D 2 A D B 3 E . A 4 C B C 5 B C . ; run; data _null_; if _n_ eq 1 then do; length order1 - order3 8 k $ 8; declare hash ha(ordered:'y'); ha.definekey('k'); ha.definedata('k','order1','order2','order3'); ha.definedone(); end; set input end=last; array o{3} order1-order3 ; array r{3} race1-race3 ; do i=1 to dim(r); if not missing(r{i}) then do; k=r{i}; rc=ha.find(); o{i}=rank; ha.replace(); call missing(of order:); end; end; if last then ha.output(dataset:'bballm'); run; proc means data=bballm noprint; output out=maxrank max=morder1 morder2 morder3 ; run; data bball; set bballm; if _n_=1 then set maxrank; array services[3] order1-order3; array maxranks[3] morder1-morder3 ; keep k order1-order3 ; do i=1 to 3 ; if services=. then services=maxranks+1; end; run; ods graphics on; proc prinqual data=bball out=tbball scores n=1 tstandard=z plots=transformations; transform untie(order1-order3); id k ; run; * Perform the Final Principal Component Analysis; proc factor nfactors=1 plots=scree; ods select factorpattern screeplot; var Torder1-Torder3 ; run; proc sort; by Prin1; run; * Display Scores on the First Principal Component; proc print; var k Prin1; run;

NKormanik · ‎05-03-2013

Beautiful. Works like a charm. Now three alternatives in hand to determine race outcome, all three yielding the same results, but apparently going about it slightly differently.

NKormanik · ‎05-03-2013

I'm sure glad you translated that mini-monster. The results are quite interesting: Runner Principal Component 1 D -2.69500 E -0.19243 A 0.75650 B 1.06547 C 1.06547 Assuming correct translation, and correct methodology, this PCA approach solves a tie issue between D and E (that resulted in a simpler form of solution script). Also, if the PC loading values have meaning (good question for ), I suppose we could say the tie between D and E wasn't even close. Whereas the performance of B and C is very close. Too, some clustering is possible, eye-balling, or using one of SAS clustering procedures. Thanks much for your help here! I still need to convert your script to accommodate my 100x100 data sets, so the work is far from over. I may have to come back and ask for more help in doing that.

NKormanik · ‎05-03-2013

Hate to keep being a bother... I've added a fifth line of data, and made two results missing. Running your script with this modified data set there appears in results nothing for runner C. Would you please check your script and change as needed? Thank you very much. Perhaps C doesn't get listed because of last place finish? data input; input rank (race1-race3) ($) ; cards; 1 D E D 2 A D B 3 E . A 4 C B C 5 B C . ; data list; set input; array race race:; do _n_ = 1 to dim(race); name = race{_n_}; output; end; drop race:; run; proc freq data=list noprint; table name*rank / sparse out=scores(drop=percent); run; proc transpose data=scores out=scoreTable(drop=_:) prefix=r_; var count; by name; id rank; run; proc sort data=scoreTable; key r_: / descending; run; proc print data=scoreTable(obs=5) obs="Final rank"; run;

NKormanik · ‎05-02-2013

Here's the stripped down version of the example mentioned above. What's needed is to translate it for the data set we have: proc means data=nicholas.bballm; output out=nicholas.maxrank max=mcsn mdurs mdurh mwas musa mspom mins mupi map mspoi; run; data nicholas.bball; set nicholas.bballm; if _n_=1 then set nicholas.maxrank; array services[10] CSN--SportsIllustrated; array maxranks[10] mcsn--mspoi; keep School CSN--SportsIllustrated; do i=1 to 10; if services=. then services=maxranks+1; end; run; ods graphics on; proc prinqual data=nicholas.bball out=nicholas.tbball scores n=1 tstandard=z plots=transformations; transform untie(CSN -- SportsIllustrated); id School; run; proc factor nfactors=1 plots=scree; ods select factorpattern screeplot; var TCSN -- TSportsIllustrated; run; proc sort; by Prin1; run; proc print; var School Prin1; run;

NKormanik · ‎05-02-2013

I'm glad for this addition by , as the actual data sets I'm dealing with are pretty large. So if the second version of transform given here is better suited to larger data sets, I'll use it. Now I need that mentioned example translated, using this present data set created.... Example 74.2 Principal Components of Basketball Rankings SAS/STAT(R) 12.1 User's Guide Thanks much!

NKormanik · ‎05-01-2013

Here is a transformed data set (thanks ), slightly expanded to include an additional runner, and some missing data (those runners missed a race). This is in line with the data format in the very similar example mentioned above by , Example 74.2 Principal Components of Basketball Rankings, SAS/STAT(R) 12.1 User's Guide. Runner Order_1 Order_2 Order_3 A 2 * 3 B 5 4 2 C 4 5 4 D 1 2 1 E 3 1 * The Basketball example is quite complicated with parts I'm not familiar with. If one of you wants to take a crack at translating that example to fit the present one, that would be fantastic. At least the data is now in the required format, a substantial step forward. I'll continue working on it as well.

Online Status	Offline
Date Last Visited	4 weeks ago

When is your SAS program good enough???

Re: Reset SAS without restarting SAS????

Reset SAS without restarting SAS????

Re: PROC GAMSELECT

Re: PROC GAMSELECT

PROC GAMSELECT

Re: Fascinating exchange with LLM

Fascinating exchange with LLM

Re: Trouble getting a macro to execute....

Trouble getting a macro to execute....

Re: When is your SAS program good enough???

Re: When is your SAS program good enough???

Re: When is your SAS program good enough???

Re: When is your SAS program good enough???

Re: When is your SAS program good enough???

Re: Using SAS Studio Online to Convert A Dataset

Re: User wants SAS code to return to Log window

Re: Logistic Regression -- Understanding Maximum Likelihood

Re: Interpreting 'Predicted' in Proc Logistic -- What's your take?

Interpreting 'Predicted' in Proc Logistic -- What's your take?

Re: Determine top five winners (Part 2) -- Breakdown of Pairs

Determine top five winners (Part 2) -- Breakdown of Pairs

Re: Transform data format to accommodate Proc expectations, Take #2

Re: Transform data format to accommodate Proc expectations, Take #2

Re: Transform data format to accommodate Proc expectations, Take #2

Re: Transform data format to accommodate Proc expectations, Take #2

Re: Transform data format to accommodate Proc expectations, Take #2

Transform data format to accommodate Proc expectations, Take #2

Re: Determine top five winners in a series of races

Re: Determine top five winners in a series of races

Re: Transform data format to accommodate Proc expectations

Re: Determine top five winners in a series of races

Re: Determine top five winners in a series of races

Re: Transform data format to accommodate Proc expectations

Re: Determine top five winners in a series of races