Date:         Wed, 10 Sep 2008 17:19:12 +0100
Reply-To:     J Sutar <jsutar@gmail.com>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         J Sutar <jsutar@gmail.com>
Subject:      Re: loop through a list of
Comments: To: "Peck, Jon" <peck@spss.com>
In-Reply-To:  <5CFEFDB5226CB54CBB4328B9563A12EE09CF6CEB@hqemail2.spss.com>
Content-Type: text/plain; charset=ISO-8859-1

Thanks Jon,

I now have the following program (see below) but as it stands it enters the
variable name in the TITLE spss command. I would like to change this so that
it enters the variable label?



BEGIN PROGRAM.
import spss, spssaux
vardict=spssaux.VariableDict(pattern=r"w12m")
for v in vardict:
 spss.Submit("""
TITLE %s

LOGISTIC REGRESSION VARIABLES  %s
  /METHOD = FSTEP(LR) age_15_24 age_25_34 age_35_44 age_45_54 age_55_64
age_65x soc_a soc_b soc_c1 soc_c2 soc_d soc_e
  /SAVE = PRED(P_%s)
  /PRINT = SUMMARY
  /CRITERIA = PIN(.05) POUT(.10) ITERATE(20) CUT(.5) .
""" % (v,v,v))
END PROGRAM.

Many Thanks

Jigs
2008/9/9 Peck, Jon <peck@spss.com>

> If you use the Python spssaux VariableDict object, you can specify a prefix
> to it and then iterate over the variables in the dictionary.  Something like
>
> begin program.
> import spss, spssaux
> vardict=spssaux.VariableDict(pattern=r"w12")
> for v in vardict:
>   spss.Submit("LOGISTIC REGRESSION VARIABLES %s /METHOD ..." % v)
> end program.
>
> HTH
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of J
> Sutar
> Sent: Tuesday, September 09, 2008 9:54 AM
> To: SPSSX-L@LISTSERV.UGA.EDU
>  Subject: [SPSSX-L] loop through a list of
>
> Hi,
>
> I would like to create program which performs similarly to the macro
> below.  I would like to adjust it such that it picks up all variables in
> the
> dataset that start with a certain prefix, for example w12m. The dependent
> variables start from w12m01 to w12m85, due to the leading zero of the first
> 9 dependent variables I struggled to create a SPSS macro. Perhaps a python
> solution would work better?
>
> Please not also that there are certain variables missing so for example
> w12m36 may not exist hence why I want it to search and find all variables
> of
> a specified prefix.
>
> Many thanks for your help.
> Jigs
>
>
>
> *
>
> http://www.spsstools.net/Syntax/RegressionRepeatedMeasure/LogisticRegressionByMacro.txt
> * I want to perform a logistic regression ten times for ten different
> variables without having to copy
> * and paste the syntax nine times.
> * Solution posted by rlevesque@videotron.ca to newsgroup on 2001/05/10.
>
> GET FILE='c:\program files\spss\employee data.sav'.
>
> * Define macro to do the job.
> */////////////////////////.
> DEFINE !logist(dv=!TOKENS(1) /iv=!CMDEND)
> LOGISTIC REGRESSION VAR=!dv
> /METHOD=ENTER !iv
> /CRITERIA PIN(.05) POUT(.10) ITERATE(20) CUT(.5) .
> !ENDDEFINE.
> */////////////////////////.
>
> * Do logistic regression of gender using salary and educ.
> !logist dv=gender iv=salary educ.
>
> * Do logistic regression of minority using salary and jobcat.
> !logist dv=gender iv=salary educ.
>
>  =====================
> To manage your subscription to SPSSX-L, send a message to
> LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD
>

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD