[R] backward prediction in multivariate multiple regression

[Ondřej Mikula]

Dear r-helpers,

I used lsfit to estimate regression coefficients from the matrices of
predictors and predicted values (i.e multivariate multiple regression
model). I wish, however, to compute rank-one matrices of predictor
values from the predicted ones.
How to do it in R? Could you suggest me appropriate function or
formula in matrix notation?
Best wishes

Ondrej Mikula
(Institue of Animal Physiology and Genetics,
Brno, Czech Republic)