Re: RES: Crosstables and order by
OK, how do you compute the "vote" then? One way, I suppose, would be to sort by the row or column total. "Clayton Santos Delfino" writes: > Hi Ben, > > Thank you for your help. > > Look the result for a table in attached file. > > I think that a possible order by to "afreq" and "dfreq" options could be > over the result computed to the "count" parameter in /cells. > > The idea is identify which entries to "B" had major e minor presence > (vote). For example, in this case, the major is Blinda. > > > Regards > > Clayton > > > > > > -Mensagem original- > De: Ben Pfaff [mailto:b...@cs.stanford.edu] > Enviada em: quinta-feira, 26 de maio de 2011 01:41 > Para: Clayton Santos Delfino > Cc: pspp-users@gnu.org > Assunto: Re: Crosstables and order by > > "Clayton Santos Delfino" writes: > >> When we are using frequencies transformations is possible to order the > >> result by dfreq and afreq but we don't have this option for >> crosstables. >> >> I think that will be a great feature we have this option in this >> transformation. >> >> What do you think about this? > > I don't know how it would work. Suppose the crosstabulation looks like > this, where the entries in cells are counts: > > A >--- > 1 2 3 > | 1 100 300 200 > B | 2 200 200 300 > | 3 300 100 100 > > In AFREQ mode, what is the correct sort order for variables A and B? > -- > Ben Pfaff > http://benpfaff.org > -- Ben Pfaff http://benpfaff.org ___ Pspp-users mailing list Pspp-users@gnu.org https://lists.gnu.org/mailman/listinfo/pspp-users
RES: RES: Crosstables and order by
I agree, the way could be to sort by the "count" parameter in column "total" with option "asc" or "desc". The vote is computed like a frequencies transformation. We have an item and verified how many times this item was cited. Again, I sending a .pdf file but with frequencies transformation. The item used is the same. Clayton -Mensagem original- De: Ben Pfaff [mailto:b...@cs.stanford.edu] Enviada em: quinta-feira, 26 de maio de 2011 11:47 Para: Clayton Santos Delfino Cc: pspp-users@gnu.org Assunto: Re: RES: Crosstables and order by OK, how do you compute the "vote" then? One way, I suppose, would be to sort by the row or column total. "Clayton Santos Delfino" writes: > Hi Ben, > > Thank you for your help. > > Look the result for a table in attached file. > > I think that a possible order by to "afreq" and "dfreq" options could > be over the result computed to the "count" parameter in /cells. > > The idea is identify which entries to "B" had major e minor presence > (vote). For example, in this case, the major is Blinda. > > > Regards > > Clayton > > > > > > -Mensagem original- > De: Ben Pfaff [mailto:b...@cs.stanford.edu] Enviada em: quinta-feira, > 26 de maio de 2011 01:41 > Para: Clayton Santos Delfino > Cc: pspp-users@gnu.org > Assunto: Re: Crosstables and order by > > "Clayton Santos Delfino" writes: > >> When we are using frequencies transformations is possible to order >> the > >> result by dfreq and afreq but we don't have this option for >> crosstables. >> >> I think that will be a great feature we have this option in this >> transformation. >> >> What do you think about this? > > I don't know how it would work. Suppose the crosstabulation looks > like this, where the entries in cells are counts: > > A >--- > 1 2 3 > | 1 100 300 200 > B | 2 200 200 300 > | 3 300 100 100 > > In AFREQ mode, what is the correct sort order for variables A and B? > -- > Ben Pfaff > http://benpfaff.org > -- Ben Pfaff http://benpfaff.org frequencies.pdf Description: frequencies.pdf ___ Pspp-users mailing list Pspp-users@gnu.org https://lists.gnu.org/mailman/listinfo/pspp-users
please join me in a prize for CTABLES
Hello friends of PSPP, I am sure I'm not the only one in need of CTABLES. I put US$100 (who should I send the check to?) . Please join me in adding to this little money pot for the PSPP people to work on CTABLES. Something that can deal with something like CTABLES /VLABELS VARIABLES=grade Group MAPReadingFall MAPReadingSpring GainInReading tGainInReading DISPLAY=LABEL /TABLE grade [C] > Group [C] BY MAPReadingFall [S][ count] + MAPReadingFall [S][ mean F40.1, STDDEV PAREN40.1] + MAPReadingSpring [S][ MEAN F40.1, STDDEV PAREN40.1] + GainInReading [s] [ mean F40.1, STDDEV PAREN40.1] + tGainInReading [s] [ mean F40.1, STDDEV PAREN40.1] /CATEGORIES VARIABLES=grade Group ORDER=A KEY=VALUE EMPTY=EXCLUDE /titles title="MAP scores in Reading, Fall to Spring, by group". or CTABLES /VLABELS VARIABLES=group MCAMPASS DISPLAY=LABEL /TABLE group [C][COUNT F40.0, ROWPCT.COUNT PCT40.1] BY MCAMPASS [C] /CATEGORIES VARIABLES=group ORDER=A KEY=VALUE EMPTY=EXCLUDE /CATEGORIES VARIABLES=MCAMPASS ORDER=A KEY=VALUE EMPTY=EXCLUDE /titles title="MCA Math pass rates by engagement & grade". And if I could add something to the wishlist, it would be to design the routine in such a way that it could be reused by the R people, if possible. ___ Pspp-users mailing list Pspp-users@gnu.org https://lists.gnu.org/mailman/listinfo/pspp-users
Re: please join me in a prize for CTABLES
Thanks for the offer. For some time, I have been asking Free Software Foundation to set up an account for PSPP to accept donations for coding. I'm starting to think they'll never act on this. The other alternative is for "us" (whoever that should be) to set up something like a PayPal account, independently of FSF, to accept donations. Though faster, this approach may suffer from an appearance of illegitimacy and lack of oversight. What do potential donors think of this idea? Set up something on our own? Or wait, possibly forever, to hear from FSF? -Jason On Thu, May 26, 2011 at 12:49:52PM -0700, Juan Zuluaga wrote: > Hello friends of PSPP, > > I am sure I'm not the only one in need of CTABLES. I put US$100 (who should I > send the check to?) . > Please join me in adding to this little money pot for the PSPP people to work > on CTABLES. > > Something that can deal with something like > > CTABLES > /VLABELS VARIABLES=grade Group MAPReadingFall MAPReadingSpring > GainInReading tGainInReading DISPLAY=LABEL > /TABLE grade [C] > Group [C] >BY MAPReadingFall [S][ count] > + MAPReadingFall [S][ mean F40.1, STDDEV PAREN40.1] > + MAPReadingSpring [S][ MEAN F40.1, STDDEV PAREN40.1] > + GainInReading [s] [ mean F40.1, STDDEV PAREN40.1] > + tGainInReading [s] [ mean F40.1, STDDEV PAREN40.1] > /CATEGORIES VARIABLES=grade Group ORDER=A KEY=VALUE EMPTY=EXCLUDE > /titles title="MAP scores in Reading, Fall to Spring, by group". > > or > > CTABLES > /VLABELS VARIABLES=group MCAMPASS DISPLAY=LABEL > /TABLE group [C][COUNT F40.0, ROWPCT.COUNT PCT40.1] BY MCAMPASS [C] > /CATEGORIES VARIABLES=group ORDER=A KEY=VALUE EMPTY=EXCLUDE > /CATEGORIES VARIABLES=MCAMPASS ORDER=A KEY=VALUE EMPTY=EXCLUDE > /titles title="MCA Math pass rates by engagement & grade". > > And if I could add something to the wishlist, it would be to design the > routine in such a way that it could be reused by the R people, if possible. > > ___ > Pspp-users mailing list > Pspp-users@gnu.org > https://lists.gnu.org/mailman/listinfo/pspp-users -- jsto...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org ___ Pspp-users mailing list Pspp-users@gnu.org https://lists.gnu.org/mailman/listinfo/pspp-users
Re: please join me in a prize for CTABLES
Jason, I think FSF or Paypal is a great idea. I teach at a university and have limited teaching enhancement funds- or could apply for - some small grants ($100? $1,000?) that might be able to provide small amounts of money for people who have the programming know-how to develop some innovations that my students would use in class (like, say, adding multicollinearity tests to OLS regression). Renan On 26-May-11 10:19 PM, Jason Stover wrote: Thanks for the offer. For some time, I have been asking Free Software Foundation to set up an account for PSPP to accept donations for coding. I'm starting to think they'll never act on this. The other alternative is for "us" (whoever that should be) to set up something like a PayPal account, independently of FSF, to accept donations. Though faster, this approach may suffer from an appearance of illegitimacy and lack of oversight. What do potential donors think of this idea? Set up something on our own? Or wait, possibly forever, to hear from FSF? -Jason On Thu, May 26, 2011 at 12:49:52PM -0700, Juan Zuluaga wrote: Hello friends of PSPP, I am sure I'm not the only one in need of CTABLES. I put US$100 (who should I send the check to?) . Please join me in adding to this little money pot for the PSPP people to work on CTABLES. Something that can deal with something like CTABLES /VLABELS VARIABLES=grade Group MAPReadingFall MAPReadingSpring GainInReading tGainInReading DISPLAY=LABEL /TABLE grade [C]> Group [C] BY MAPReadingFall [S][ count] + MAPReadingFall [S][ mean F40.1, STDDEV PAREN40.1] + MAPReadingSpring [S][ MEAN F40.1, STDDEV PAREN40.1] + GainInReading [s] [ mean F40.1, STDDEV PAREN40.1] + tGainInReading [s] [ mean F40.1, STDDEV PAREN40.1] /CATEGORIES VARIABLES=grade Group ORDER=A KEY=VALUE EMPTY=EXCLUDE /titles title="MAP scores in Reading, Fall to Spring, by group". or CTABLES /VLABELS VARIABLES=group MCAMPASS DISPLAY=LABEL /TABLE group [C][COUNT F40.0, ROWPCT.COUNT PCT40.1] BY MCAMPASS [C] /CATEGORIES VARIABLES=group ORDER=A KEY=VALUE EMPTY=EXCLUDE /CATEGORIES VARIABLES=MCAMPASS ORDER=A KEY=VALUE EMPTY=EXCLUDE /titles title="MCA Math pass rates by engagement& grade". And if I could add something to the wishlist, it would be to design the routine in such a way that it could be reused by the R people, if possible. ___ Pspp-users mailing list Pspp-users@gnu.org https://lists.gnu.org/mailman/listinfo/pspp-users -- Renan Levine Department of Political Science University of Toronto - Scarborough renan.lev...@utoronto.ca http://individual.utoronto.ca/renan (416) 208-2651 ___ Pspp-users mailing list Pspp-users@gnu.org https://lists.gnu.org/mailman/listinfo/pspp-users
