Output discrepancy between SAS and R -
i have coded same program in r , in sas (university edition running on oracle virtualbox on mac) , i'm noticing discrepancy - means messed up.
the first discrepancy appears in principal component analysis run , believe has options intrinsic functions available in 2 programs. diligently working on looking on documentation grateful assistance of brilliant experts on site.
i have been using file - https://drive.google.com/file/d/0b9oqam9ykac3bepicestrw8wuzg/view?usp=sharing - test 2 programs.
in r, code pca simple:
pca1=prcomp(predictor_matrix, scale.=true)
and first observation transformed following
<1.01, -0.79, -0.03, -1.08, 1.86, -0.13, 0.04, -0.03, 0.02, -0.01>
in sas, code pca looks like:
proc factor data=regressors simple corr mineigen=0 /*retain eigenfunctions*/ nfactors=10 /*print optional data*/ out=newdata /*get transformed data*/ ; run;
and first observation transformed following
<0.55, -0.53, -0.02, -0.89, 1.96, 1.45, 0.89 , 1.18, 1.15, 1.46>
now - eigenvector matrices identical in both programs eigenvalues different, i'm thinking problem has scaling. brand new sas , use pointers on how results of these 2 programs converge.
the sas proc pca proc princomp
, not proc factor
. try
proc princomp data=regressors n=10 out=newdata; run;
Comments
Post a Comment