Re: CTABLES

[Alan Mead]

What does CTABLES do?  And what aspects of CTABLES would be critical for users?  I think a plan should be the first step. If CTABLES were part of PSPP, is the output OK?  Or would that be the next hurdle to using PSPP that the output isn't the same as SPSS (e.g., I find it harder to copy-and-paste from PSPP into LibreOffice than SPSS; when you copy a chart from SPSS for Windows, it's actually placing several versions of the chart in the paste buffer). If the output won't ultimately be useable, that makes adding CTABLES a waste of time...

Also, I'm sure CTABLES is very important to some people but a lot of SPSS users probably have never used CTABLES (just like they've never used any of the other special modules SPSS publishes or even the new features like python scripting). I think that contributes to its absence in PSPP.

Finally, I know how statistics are calculated, but not only don't I know what CTABLES does but I understand that the emphasis is on simply tallies across complex breakdowns... I don't have any feel for how one efficiently does this in code. 

BTW, I think a better big project would be to enable PSPP to read the SPSS output format.  I know Ben enjoys reverse engineering things. Thanks largely to Ben and PSPP contributors, the SPSS file format is widely read (e.g., the R foreign package is based on Ben's code to read SPSS .SAV files) but the output is completely opaque to all programs except SPSS. In fact, it's worse than that because there are versions of SPSS that won't read other versions.  That's actually a huge problem for people who used SPSS to analyze data and all they kept was the output files or anyone who ever saved the output in the hope that they could read it years later....

-Alan

On 11/7/2015 5:22 AM, Matthias Faeth wrote:

I think we have to find somebody who has done that kind of project steering before (I haven't). Can everybody look into their contacts to find somebody who would have the skills and will to steer such a project?

However I'd volunteer to participate in the workgroup defining the priority of the functions of the CTABLES command. Maybe Frans and Alan would want to join?

I agree with John that a commercial implementation would produce 5 figured cost. I've already talked to a programmer on this and he came up with the same estimate. So doing it commercially seems to be no way forward.

Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
e: address@hidden

2015-11-07 7:19 GMT+01:00 John Darrington <address@hidden>:

Funding the implementation of a fully blown CTABLES at commercial rates,
I think would cost well into 5 figures.

However, it is perhaps possible, that a smaller subset of CTABLES could
be undertaken by an altruistic programmer who has some time on his
hands and wishes to earn a bit of money for a few months.

The problem, is (and this is where big companies have the advantage) most
programmers are not good administrators.  To crowdfund such an exercise
one would first have to agree on which subset of CTABLES to undertake.
That means, a detailed specification, with examples showing exactly what
would be expected, which features of CTABLES should be included, and perhaps
more importantly, which are not expected.  Once that has been agreed,
an estimate of the time and cost could be prepared an appeal for funds could begin.
This kind of organisation, takes skill and effort, and would probably
have no tangible reward, unless some percentage of the total was agreed
beforehand.

J'


--
Avoid eavesdropping.  Send strong encryted email.
PGP Public key ID: 1024D/2DE827B3
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://sks-keyservers.net or any PGP keyserver for public key.


_______________________________________________
Pspp-users mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/pspp-users

_______________________________________________
Pspp-users mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/pspp-users

-- 

Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.

science + technology = better workers

+815.588.3846 (Office)
+267.334.4143 (Mobile)

http://www.alanmead.org

Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing: http://www.iacat.org/jcat