R help - Jan 2004 - regression to get non-negative parameters

Hi, 
I am working on a model sort of like a credit assignment  problem. 
The goal is to get non-negative weights for each input variable and 
discount the predictors (down to 0) that lead to no credits.

I am trying out the lm fitting and find out they give a lot of negative weights.
I am totally stuck now.  

Is there any reasonable fitting function I can use to get non-negative weights?


Please advise.

thanks a lot,


Yun-Fang


	[[alternative HTML version deleted]]