On Mon, Jan 16, 2006 at 05:33:53PM -0800, Ben Pfaff wrote: > Jason Stover <[EMAIL PROTECTED]> writes: > > > How do you all feel about including a time series procedure in PSPP? > > It would be one or more of TSMODEL, AREG and ARIMA. > > > > Specifically, I was thinking of adding a procedure to fit ARMA models. > > I'm in favor of adding any procedure that is likely to be of use > to our users and developers, especially when they happen to > correspond to procedures in SPSS (but even when they don't). > > Is REGRESSION in good shape? (Besides the need for temporary > files, which I am working on.) It will be a nice new feature to > announce in our next release.
What's there is ready for testing. It produces the most essential output: An ANOVA table, a table of coefficients and standard errors, as well as the table with R squared and the covariance matrix. I think the most important task it still cannot do is save the residuals to the data. -Jason -- [EMAIL PROTECTED] SDF Public Access UNIX System - http://sdf.lonestar.org _______________________________________________ pspp-dev mailing list [email protected] http://lists.gnu.org/mailman/listinfo/pspp-dev
