Hi, I work in a kitchen production factory and I try to predict sales of kitchen for an horizon of 12 weeks, and I have to turn into account promotions. My sales are express in number of command, my promotion are express with dummy variables. (1 if promotion, 0 else). The first problem is that my time serie contain a trend and a seasonality. (but the serie is stationary) I use the VAR function from the package vars, but after red the package's guide, I'm not sure if this function could estimate a model for this kind of serie (trend + seasonality), and it is important for the forecast. I also use the function estVARXls from the package dse but the results are strange. The second problem is that I have to use the future promotion to make the forecast; the var function can give me a forecast for the sales, but I don't know if we can seize the future promotions (that I know and which I have) in the function (predict(), or another one if it exists) to obtain forecasts according these promotions and not those obtain by the predict() function. Could somebody help me about these 2 questions? just ask me if you need more information Thank you by advance, M. -- View this message in context: http://r.789695.n4.nabble.com/prediction-of-sales-with-VAR-model-tp4633107.html Sent from the R help mailing list archive at Nabble.com.