Date:         Sat, 3 Sep 2005 11:07:52 -0400
Reply-To:     Raynald Levesque <rlevesque@videotron.ca>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Raynald Levesque <rlevesque@videotron.ca>
Subject:      Re: Automated use of data in output file?
In-Reply-To:  <bfed04db050902213129afa7e8@mail.gmail.com>
Content-type: text/plain; charset=Windows-1252

Hi

For some examples see http://www.spsstools.net/SampleSyntax.htm#Bootstrap

If you need to read data from Pivot Tables in the spo, OMS is probably
something to look at.

Regards

Raynald Levesque Raynald@spsstools.net
Visit my SPSS site: http://www.spsstools.net


-----Original Message-----
From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU]On Behalf Of
Craig Johnson
Sent: September 3, 2005 12:32 AM
To: SPSSX-L@LISTSERV.UGA.EDU
Subject: Automated use of data in output file?


I'm doing some Monte Carlo work and would like to be able to streamline the
process some. I have written syntax to create a normal data set, set the
mean and standard deviation and then round the data. I'm looking for a way
to basically have SPSS read the output file (such as mean, percentiles etc.)
in such a way that I can plug them back into a formula in syntax and
continue on in a sort of cascading method. That way I can run lots of
variations on the experiment with only changing a few lines of code NOT
spending hours doing the exact same thing entering manual information. I'm
not sure if this is possible but I'm hopeful seeming it is possible to do
something similar in data base programs……..this whole output file that does
not have a real "data base" structure that I can pull information back from
a results variable has me some what stumped. Even if I could figure out a
method to save results via a matrix that SPSS could turn around and use
would be good. (Assuming one could create a macro, script or piece of syntax
that would save and import all in one click of the mouse.)

 For example:

 1) Create data set and create output file

2) Save output file to default location. (Including make new folder if
needed.)

3) Pause in program.

4) Import information from output file.

5) Pause in program

6) Complete calculations based on the information found in imputed file.

7) "Update" output file.

8) Done!

 I would like to be able to do these things:

 1) Create Data set (Done.)

2) Run mathematical computation – show results in output file

3) Run exploratory statistics to calculate outliers/percentiles

4) Take the calculated percentiles shown in the output and plug them back
into a calculate formula to select out outliers via Tukey's hinges.

5) Create "Select If" statement based on Tukey's hinges calculation

6) Run mathematical computation – show results in output file

7) I can then compare the before and after to see the results of the
outliers.

 Any help on this would be appreciated. I hope there is something like this
out there. If there isn't there should be. =) What good are results if you
have to manually play with them every time? =)

 Thanks!

 Craig