Recommendation on error messages in v18

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Recommendation on error messages in v18

Dennis Deck
Years ago when SPSS output was a simple text file it was very easy and fast to search on the word “ERROR” to check for error messages.
 
Now (in v18) the Output file can get very large and error messages are very hard to see.
Doing a text search for “>error” in Output can be very slow on all but the simplest job.
The color coding in the syntax file is of some help in avoiding errors but it primarily helps catch typos in the commands themselves.
 
It seems to me that a much more efficient process could be made available for this purpose. 
For example, why couldn’t one select an option to only display error messages — that would be a first step to ensure the job ran cleanly.
Error messages could be tagged in some way so one could search for the next one quickly.
Dennis Deck, PhD
RMC Research Corporation
111 SW Columbia Street, Suite 1200
Portland, Oregon 97201-5843
voice: 503-223-8248 x715
voice: 800-788-1887 x715
fax:  503-223-8399
[hidden email]
 
 
 
Reply | Threaded
Open this post in threaded view
|

Re: Recommendation on error messages in v18

Jon K Peck

Finding error message can indeed be a chore even though they are now displayed in a syntax window pane if you run your syntax that way.

Here is one way to make these easier to catch.
oms select tables logs /if subtypes='Warnings'
destination outfile="c:/temp/warnings.txt" format=text.
<lots of syntax>
omsend.

warnings.txt will contain all the warning blocks and the log blocks, so they can be searched easily in any text editor.

HTH,

Jon Peck
SPSS, an IBM Company
[hidden email]
312-651-3435



From: Dennis Deck <[hidden email]>
To: [hidden email]
Date: 04/08/2010 05:28 PM
Subject: [SPSSX-L] Recommendation on error messages in v18
Sent by: "SPSSX(r) Discussion" <[hidden email]>





Years ago when SPSS output was a simple text file it was very easy and fast to search on the word “ERROR” to check for error messages.
 
Now (in v18) the Output file can get very large and error messages are very hard to see.
Doing a text search for “>error” in Output can be very slow on all but the simplest job.
The color coding in the syntax file is of some help in avoiding errors but it primarily helps catch typos in the commands themselves.
 
It seems to me that a much more efficient process could be made available for this purpose.  
For example, why couldn’t one select an option to only display error messages — that would be a first step to ensure the job ran cleanly.
Error messages could be tagged in some way so one could search for the next one quickly.
Dennis Deck, PhD
RMC Research Corporation
111 SW Columbia Street, Suite 1200
Portland, Oregon 97201-5843
voice: 503-223-8248 x715
voice: 800-788-1887 x715
fax:  503-223-8399

ddeck@...
 
 
 

Reply | Threaded
Open this post in threaded view
|

Re: Recommendation on error messages in v18

Luca Meyer-3
In reply to this post by Dennis Deck
<base href="x-msg://270/">
Hi Dennis,
Have you updated the most recent PASW patch?
I myself experienced and reported to SPSS very long waiting time while I was running PASW 18.0.0 on MAC OS while searching for "ERROR" in the output. Now that I am using 18.0.1 the time has become much more acceptable.
Luca

Luca Meyer
www.lucameyer.com
PASW Statistics v. 18.0.1 (13-nov-2009)
Mac OS X 10.6.3 (10D573) - kernel Darwin 10.3.0

Il giorno 09/apr/2010, alle ore 01.25, Dennis Deck ha scritto:

Years ago when SPSS output was a simple text file it was very easy and fast to search on the word “ERROR” to check for error messages.
 
Now (in v18) the Output file can get very large and error messages are very hard to see.
Doing a text search for “>error” in Output can be very slow on all but the simplest job.
The color coding in the syntax file is of some help in avoiding errors but it primarily helps catch typos in the commands themselves.
 
It seems to me that a much more efficient process could be made available for this purpose. 
For example, why couldn’t one select an option to only display error messages — that would be a first step to ensure the job ran cleanly.
Error messages could be tagged in some way so one could search for the next one quickly.
Dennis Deck, PhD
RMC Research Corporation
111 SW Columbia Street, Suite 1200
Portland, Oregon 97201-5843
voice: 503-223-8248 x715
voice: 800-788-1887 x715
fax:  503-223-8399
[hidden email]
 
 
 







Reply | Threaded
Open this post in threaded view
|

Re: Recommendation on error messages in v18

Gary Oliver-3
<BASE href=x-msg://270/>
Colleagues at SPSS
 
Nothwithstanding patches, I think Dennis' suggestion is one which should appeal to SPSS on the grounds of rigor and diagnostic control.
 
It seems to me that output which contained the string error could be parsed and filtered to a separate file dedicated to error listings. I know I for one would find this both helpful and comforting.
 
Warm regards/gary


From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Luca Meyer
Sent: Friday, April 09, 2010 5:28 PM
To: [hidden email]
Subject: Re: Recommendation on error messages in v18

Hi Dennis,
Have you updated the most recent PASW patch?
I myself experienced and reported to SPSS very long waiting time while I was running PASW 18.0.0 on MAC OS while searching for "ERROR" in the output. Now that I am using 18.0.1 the time has become much more acceptable.
Luca

Luca Meyer
www.lucameyer.com
PASW Statistics v. 18.0.1 (13-nov-2009)
Mac OS X 10.6.3 (10D573) - kernel Darwin 10.3.0

Il giorno 09/apr/2010, alle ore 01.25, Dennis Deck ha scritto:

Years ago when SPSS output was a simple text file it was very easy and fast to search on the word “ERROR” to check for error messages.
 
Now (in v18) the Output file can get very large and error messages are very hard to see.
Doing a text search for “>error” in Output can be very slow on all but the simplest job.
The color coding in the syntax file is of some help in avoiding errors but it primarily helps catch typos in the commands themselves.
 
It seems to me that a much more efficient process could be made available for this purpose. 
For example, why couldn’t one select an option to only display error messages — that would be a first step to ensure the job ran cleanly.
Error messages could be tagged in some way so one could search for the next one quickly.
Dennis Deck, PhD
RMC Research Corporation
111 SW Columbia Street, Suite 1200
Portland, Oregon 97201-5843
voice: 503-223-8248 x715
voice: 800-788-1887 x715
fax:  503-223-8399
[hidden email]