Hi Jack,
I do not know the theory for Correspondence Analysis but I gave a try to
your package.
It failed to produce the plot because it used decimal commas instead of
decimal points. (You did that on purpose, didn't you? ;) )
After correcting the numbers the plot was produced correctly (only one).
Off course, running in English produce the four plots.
I am not complaining, just letting you know :).
Hélio
On Thu, Feb 5, 2015 at 5:08 PM, Riccardo (Jack) Lucchetti <
r.lucchetti(a)univpm.it> wrote:
This is both an announcement and a plea for help: some time ago, I put
together some functions for simple correspondence analysis, which I am not
at all an expert of (I used to be pretty good at this stuff in the late
1980s, then I stopped following the literature).
These functions are now available as a function package, that you'll find
on the server under the "corresp" name. I decided to make this stuff public
because it's a nice illustration of how to use string arrays, so you'll
need a pretty recent version of gretl to run it.
Besides, in order to run the example, you'll also need the datafile
attached here, which contains a classic example, used in practically all
software implementations as a simple benchmark (see Greenacre (2007),
"CORRESPONDENCE ANALYSIS in PRACTICE", Chapman & Hall).
The plea for help is: is there anybody out there who will adopt the
package? I'm not at all sure I implementend things the right way and, as I
said, I haven't followed the literature for a looooong time (and I have no
intention/time for doing so).
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------
_______________________________________________
Gretl-devel mailing list
Gretl-devel(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-devel