Date: Wed, 12 Oct 2005 10:46:00 +0200
Reply-To: MOSER Winfried <email@example.com>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: MOSER Winfried <firstname.lastname@example.org>
Subject: AW: Clementine: properties of varibles, export of models
Content-Type: text/plain; charset="iso-8859-1"
by now i am not even an beginner in clementine, but next week (monday and tuesday) i will visit an spss-course about clementine in munich. i will ask for you if you want me to.
Von: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] Im Auftrag von Spousta Jan
Gesendet: Mittwoch, 12. Oktober 2005 10:40
Betreff: Clementine: properties of varibles, export of models
Hi all Clementine users,
I spent now two days in Vienna and spoke with a colleague SPSS user who will buy a data mining tool and decides between Clementine (SPSS) which he tries just now, and Enterprise Miner (SAS) which he will try in the next future. So we exchanged experiences: I told him the horror stories about SAS EM and he criticized Clementine. :-) He told me, that his two biggest problems with Clementine are:
1) All variables in the file should have one of six measurement levels. The setting is not easily copied from the source SPSS files. If he will work with 10 variables in a file with 1000 variables, he must define the measurement levels for all 1000 variables and not only the 10 which he needs.
2) It is not possible to export the final scoring syntax into the SPSS language - the only possibility other than Clem is C (if I remember it
properly) which is not convenient for him.
Since I do not know Clementine, my question is whether it is both true or whether there is a workaround for the two issues. The colleague is of course willing to change to SAS, but it would mean a lot of work for him, too.