pspp-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Running Production Job on PSPP


From: Matthew Silver
Subject: Re: Running Production Job on PSPP
Date: Mon, 24 Jun 2013 15:16:16 -0400

Since I'm not experienced with either SPSS or PSPP there might already be a solution that I missed. But this might clarify things further:

The file validation_production_job.spj, which is naturally in XML, contains instructions for which .sps files to run. It also contains a segment indicating where to output a view file (a log of the actions executed for each .sps file that can be in the format of an spss view [.spv], word doc, excel spreadsheet, or pdf).  In order to run a .spj file called “%TEMP%” (which I have defined earlier in the batch file) that uses data from the file %DATA%, I type

"C:\Program Files\IBM\SPSS\Statistics\21\stats.exe" %TEMP% -production silent -symbol @datafile %DATA%

Someone suggested that the equivalent statement in PSPP would be among the lines of "pspp syntax1 & pspp syntax2" but I'm not sure how to deal with things like -production or silent or -symbol or @datafile ...



On Mon, Jun 24, 2013 at 12:40 PM, John Darrington <address@hidden> wrote:
What that page doesn't explain, is what can a "production job" do, that can't be
done with parallel executions of SPSS/PSPP.

Ok,  PSPP cannot run two syntax files concurrently.  But, unless you have a multi-core
or multi-processor machine, doing so will save little if any time.  If you do have
a multi-processor machine, or perhaps a cluster, then what can a "production job" do
that isn't done by running "pspp syntax1 & pspp syntax2" ?

The web page also talks about being able to run processes on remote servers, but
presumably some program/service has to be installed there - so why not just install
PSPP on the remote server and run that?

The only possible advantage I can see, is if a "production job" can somehow collate
and aggregate the results of two or more parallel jobs.  But the page makes no
mention of such an ability.

It seems that it's primary selling point is
"Production jobs are useful if you often run the same set of time-consuming
analyses, such as weekly reports."
But again, I don't see why you need anything special to do that. Simply write a
syntax file to analyse the report, put it in a well know file, and set your
crontab (or equivalient under windows) to run this syntax each week.


Am I missing something?


John


On Mon, Jun 24, 2013 at 10:24:47AM -0400, Matthew Silver wrote:
     The .spj file is a production job which runs several .sps files at once. A
     more detailed explanation can be found at the link

     http://publib.boulder.ibm.com/infocenter/spssstat/v20r0m0/index.jsp?topic=%2Fcom.ibm.spss.statistics.help%2Fidh_prod.htm


--
PGP Public key ID: 1024D/2DE827B3
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://keys.gnupg.net or any PGP keyserver for public key.


Attachment: be_sd_cust_step1.sps
Description: Binary data

Attachment: be_sd_cust_step2.sps
Description: Binary data

Attachment: validation_diff.sps
Description: Binary data

Attachment: validation_extract_dla_1302_main.sps
Description: Binary data

Attachment: validation_production_job.spj
Description: Binary data


reply via email to

[Prev in Thread] Current Thread [Next in Thread]