Hi everybody, I read the 1st part of Udo's blog about using external models in SAS Forecasting. Has anyone tried any external models in practice? http://blogs.sas.com/content/forecasting/2011/09/19/guest-blogger-udo-sglavo-on-including-r-models-in-sas-forecast-server-part-1-of-2/ Have a good weekend! Wevergton ... View more

Hello folks, anyone know hot to use the eventby in hpfdignose? I'm in doubt about the layout of data. Thanks. ... View more

Hello everyone! I know a daily forecast isn't easy thingh to do, but I belive is possible give a reasonable solution using SAS High Performance Forecasting in supply chain demand for a short shelf-life products. For a "regular" period the traditional time series metrics doing very well, but for a period before the holiday, for example, the prediction wasn't so satisfatory. I believe a hybrid model ARIMA+fuzzy model should reduce the variability of the demand information. The question is: Have anyone tried combine time series models and fuzzy or network models in SAS ? Thanks. ... View more

Hello everyone ! I'm using SAS High Performance Forecasting 2.3 and I have a question about accumulation step. Is there any option to ignore time series with no data recorded in the last 3 months, for example ? In mostly cases after 3 months missing values represents no activity in production line. Thanks ! ... View more

Hello everyone ! It's my very first time in this forum. I'm trying use 2 prices as an exogenous in a ARIMAX model. But I need to define a restriction in the first price (because doesn't make sense increase the price and the sales increase to), and leave the second price (parameter) free. I know it's possible apply a restriction (positive or negative) on input, but i don't know if it's possible define different rules for different inputs. ... View more