st: Non negative dependent variable

[Laurie Molina <molinalaurie@gmail.com>]

Dear statlisters,
I am triying to imput rental prices for houses to a sample using some
characteristics of the construction which are shared between my data
and the data in another database in which rental prices declared.
When i use OLS i get negative fitted values both in the original
database as in my database, and i don't want it to happen because i am
doing the regresion for predictive purpouses.

I have to question i hope you could help me with:
1. How can ensure that the predicted values will be non negative?
2. Should i use all the possible variables in my regression or only
the ones that are statistically significant? (I ahve a predictive, not
explanatory purpose).

Thank you very much in advance.

Laurie
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/