[R] reconstruction of data after PCA and interpolation

[Poizot Emmanuel]

Dear all,

I have a matrix of size N x M. I purchase a PCA analysis through prcomp 
function. Then I keep the H eigenvectors which explain 90% of the total 
variance and interpolate each vectors of the matrix H x M, to obtain a 
new matrix of size H x K (K > M). My question is : from this last matrix 
(H x K), how to (re)construct a matrix of size N x K which have the same 
units of the inital matrix (N x M).
Regards

-- 
Cordialement

------------------------------------------------
Emmanuel Poizot
Cnam/Intechmer
B.P. 324
50103 Cherbourg Cedex

Phone (Direct) : (00 33)(0)233887342
Fax : (00 33)(0)233887339
------------------------------------------------