Hi Ben Thanks for the comprehensive explanation why CTABLES depend on the ability to read SPV files. Do I understand correctly that this dependency comes only from the pivot-function in the SPV file - meaning that you doubleclick on a table to change pages?
I mean by that the following: The syntax of CTABLES is: "CTABLES downvariable BY acrossvariable BY pagevariable" Pages are shown in the SPV file only by doubleclicking and choosing another page. The function is a pain anyway because it is uncomfortable and also (at least for me) not working under all circumstances. So I can confirm Alans arguments on this although I have no idea why rendering a new page is not always working. To come back to the topic of implementing CTABLES: 1. Can we avoid the dependency on understanding SPV if we strip the CTABLES command of the page function? So to support only 2 dimensions in the output? This is probably what most users need. 2. Or do you think that is not worth to do, because once the SPV issue is solved the whole CTABLES output would have to be reprogrammed? Or are there other functions of the CTABLES command (like plots or graphs) that depend on the SPV issue, too? 3. As I do not have any knowledge about TABLES: does TABLES have the same dependencies on pivot and SPV? Would it be a solution to come back to the proposal of ftr and Frans to opt for TABLES instead of CTABLES? Matthias Fäth Im Mediapark 12 50670 Köln t: 0221-2907973 m: 0171-9832175 e: m.fa...@gmx.de 2015-11-08 23:54 GMT+01:00 Ben Pfaff <b...@cs.stanford.edu>: > On Sun, Nov 08, 2015 at 06:20:25PM +0100, Matthias Faeth wrote: > > I'm not sure if I get the discussion under the label CTABLES here: why > > should "production jobs" replace CTABLES? > > > > Or is it just that there are 3 open topics by now > > 1. CTABLES > > 2. Production jobs > > 3. Reading .spv files > > > > and the discussion is on which to focus time and energy? > > Production jobs are a separate topic, and John has changed the subject > for the emails that focus on production jobs (thanks, John). > > SPV files have more to do with CTABLES than you may realize. CTABLES is > a heavy-duty application of pivot tables. It relies on a lot of > features that aren't well documented for SPSS; the various SPSS > explanations of pivot tables are terribly vague. The SPV file format, > however, is a pretty clear specification for how SPSS pivot tables > really work, and once I came to learn about it in detail I understood > much better how output formatting really works. So this is an essential > building block for CTABLES and will be important for eventually > implementing it. >
_______________________________________________ Pspp-users mailing list Pspp-users@gnu.org https://lists.gnu.org/mailman/listinfo/pspp-users
