# Prediction/time series/data problem

I have a poverty dataset with rows containing different places and different years and I have to predict the poverty percentage for those places next year.

Overall I have over a thousand rows. But I don't have enough data points for each of the places. Actually, the max data point I have for each of the places is only 2.

Obviously, time series isn't very likely. I tried linear interpolation to increase my data points but that wasn't very reliable. I'm also looking into imputation.

Right now, I'm looking into predict() of R, but I'm not sure that can actually give me anything that will also consider the year.

Any suggestions as to how to deal with this problem?

• stats::predict is a generic function, it doesn't do any actual prediction by itself. You pass it a model object and it will compute the prediction for that model. If you pass in a time series model it will "consider the year". This isn't so much a question about R, you just need to find a good model for this situation (many related, very short time series) which will allow you to share information about the dynamics across series. – Chris Haug Feb 27 '17 at 15:03