Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "Nick Cox" <n.j.cox@durham.ac.uk> |
To | <statalist@hsphsun2.harvard.edu> |
Subject | st: RE: exponential smoothing |
Date | Mon, 16 Aug 2010 18:58:17 +0100 |
Predicting the logarithm of sales and back-transforming is one possibility. Another is that your sales are fluctuating too erratically to be predictable by this method. Nick n.j.cox@durham.ac.uk Schöler, Lisa I want to do exponential smoothing for sales with Stata. I used the command .tssmooth dexponential sales = ussalesdols000, forecast(1) Sometimes I get negative out of sample forecasts which can't be true for sales. Is there a way to include a constraint so the out of sample forecasts will not get negative? * * 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/