[Gretl-devel] Survey Reweighting

Hi, I've been using Gretl with my students at the Open University in the UK for three years not and they and I really like using it, so thank you all. I was thinking of something I could give back. Would you be interested in an implementation of the Survey reweighting algorithms described in (e.g): Creedy 2003 http://www.treasury.govt.nz/publications/research-policy/wp/2003/03-17/tw... and Calibration Estimators in Survey Sampling Jean-Claude Deville and Carl-Erik Sarndal Journal of the American Statistical Association , Vol. 87, No. 418 (Jun., 1992), pp. 376-382 These generate a set of sample weights for a sample survey that guarantee that the grossed up survey hits certain targets (numbers of males and females, numbers in different types of employment, etc.) subject to the weights being as close as possible (in some sense) to some initial set of weights. As some of you may know, these techniques go back a long way under various names; in my world they are often simply referred to as 'Calmar' after the SAS program: http://www.restore.ac.uk/PEAS/saspackage.php) I have an implementation of this written in Ada (I'm Old Skool); it's not derived from the SAS version and (I think) uses different algorithms, taken from the Deville paper. I've been looking at the Gretl code and I think I can see roughly how to port it. I'm only slowly getting to grips with the Gretl code, so would need a certain amount of hand-holding, I'd expect, especially on the user-interface. Is that something that would interest you? I realise it's not strictly econometrics, but this is something widely used in Microsimulation modelling. thanks, Graham -- Graham Stark, Virtual Worlds Research http://www.virtual-worlds-research.com 136 Hainault Avenue, Milton Keynes, MK14 5PG, UK t: (+044) 01908 618239 skype: graham_k_stark m:(+044) 075283 45116