On Sat, March 4, 2006 07:58, Al Leong wrote: Thank you, I didn't know about grocer. I'll have a look when I have time (not today!!!). However, it seems that the list of things we could take from grocer is not so big after all. From the project web page: - ordinary least squares and various single equation methods (autocorelated models, instrumental variables, non linear least squares, limited dependent variables, robust methods,...) we've got all of these - specification tests (multicolinearity, autocorelation, heteroskedasticity, normality, predictive failure,...) ditto - simultaneous equations methods (SUR, two and three stage least squares,...) ditto, plus liml and fiml - VAR, VECM, VARMA and GARCH estimation VECM needs some more love, but that's on the todo list. We haven't got varma. However, you'll agree that VARMAs are not the most widely used models in econometrics :-) Still, Allin has recently committed code that provides the foundation for VARMA estimation should we decide we need it. - the Kalman filter and time varying parameters estimation see VARMA - unit root tests (ADF, KPSS,...) and cointegration methods (CADF, Johansen,...) we've got these too - various business cycle tools: HP, Baxter-King and Christiano-Fitzgerald filters, the Bry-Boschan-Harding-Pagan procedure for the datation of turning points, spectral analysis We've haven't got Christiano-Fitzgerald, which is a simple modification to Baxter-King and should be easy to do if we need it. I don't know what the Bry-Boschan-Harding-Pagan procedure is, I'll take a look at it. The spectral analysis tools we have as of now are functional if a bit basic; again, we may want to make some progress here (eg. cross-spectra) - basic panel data estimation: fixed and random effects, between estimation We've got these. More to come in the future. - numerous time series disaggregation methods: Chow-Lin, Litterman,... None of these, but they should be easy to "script out". # some rare -and useful- ones: - a pc-gets like device, that performs automatic general to specific estimations Personally, I'd rather avoid this. With all due respect for David Hendry & co., I'm a little wary of the magic machine where you put data in and the "right" model miraculously comes out. But maybe it's just me becoming an old fart with age. - a contributions device, that provides contributions of exogenous variables to an endogenous one for any dynamic equation What's this? -- Riccardo "Jack" Lucchetti Dipartimento di Economia Facoltà di Economia "G. Fuà" Ancona