13.07.2015 Views

View - Kowalewski, M. - Virginia Tech

View - Kowalewski, M. - Virginia Tech

View - Kowalewski, M. - Virginia Tech

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

PALEONTOLOGICAL SOCIETY PAPERS, V. 8, 2002nrow(Y2)))*(180/&pi);C=sum(COS(Y))/nrow(Y); S=sum(SIN(Y))/nrow(Y);R=sqrt(C**2+S**2)*nrow(Y); mean_a=arcos(C/(R/nrow(Y)))*(180/&pi);Fstat=((nrow(Y)-2)*(R1+R2-R))/(nrow(Y)-R1-R2);F=a1||a2||mean_a||Fstat; FINISH WATSON;START BOOT(X1,X2,times,dist);RUN watson(X1,X2,aF); X=X1//X2; k=nrow(X1);j=nrow(X2);RUN mixup(X,times,template);Do i=1 to times; Y1=X[template[1:k,i]];Y2=X[template[(k+1):(k+j),i]];RUN watson(Y1,Y2,F); rF=rF//(i||F); END; rand=rF;act=shape(aF,nrow(rand),ncol(aF)); dist=rand||act; FINISH BOOT;RUN BOOT(X1,X2,&times,dist);CREATE OUT from DIST [colname={‘i’‘r1’‘r2’‘mean-r’‘rF’‘a1’ ‘a2’ ‘mean_a’ ‘aF’}];APPEND from DIST; CLOSE OUT;DATA report; set OUT; if i=1; keep a1 a2 mean_a aF;DATA count; set OUT; if rF>=aF then p=1; else p=0;PROC univariate noprint; var p; output out=last sum=s N=n;DATA prob; set last; n=n+1; p=(s+1)/n; keep p n;DATA final; merge prob report;PROC print data=final noobs split=’*’;label a1=’mean angle for the first sample’; label a2=’meanangle for the second sample’; label mean_a=’mean angle for pooleddata’; label aF=’Waston-Williams Stat. without K-factor correction’;label n=’number of random samples (# iterations + 1)’; labelp=’probabil. that 2 samples have the same mean angle’; RUN; QUIT42

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!