SERIOUS problem with Multiple datasets in MATRIX procedure

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

SERIOUS problem with Multiple datasets in MATRIX procedure

David Marso
Administrator
The simplest most OBVIOUS code aborts with a truly inexplicable/inexcusable error!
This came up awhile ago, but no acceptable solution has been posted:
One "solution" which "works" but is unacceptable for any code running on machines where the user does not have administrative rights was posted.
Yes, running as admin 'fixes' the problem, but the solutions I and many others provide to clients
CANNOT depend or REQUIRE that the user has admin rights.  
Nor should we have to document right click>run as administrator prior to starting SPSS.
What is the point of supporting multiple datasets if they cannot be reliably used across ALL procedures.
I know another solution is to write to disk, But I really HATE writing ANYTHING to a users disk unless it is absolutely necessary.  Besides, I have important code which WORKED in ver 21 without Admin rights.
I hope this will be fixed in 22.01 or .??  How did this ever pass QA?
Any ideas of what is actually going on and the nature of the obviously incorrect error message?
----
DATA LIST LIST / a b c.
BEGIN DATA.
1 2 3
3 3 4
2 6 5
END DATA.
DATASET NAME abc.
DATASET DECLARE results .
DATASET ACTIVATE abc.
DATASET DISPLAY.

MATRIX.
GET abc / FILE * .
COMPUTE Tabc = T(abc).
SAVE Tabc / OUTFILE results .
END MATRIX.
DATASET DISPLAY.

Run MATRIX procedure:
 
>Error # 34 in column 21.  Text: results
>SPSS Statistics cannot access a file with the given file specification.  The
>file specification is either syntactically invalid, specifies an invalid
>drive, specifies a protected directory, specifies a protected file, or
>specifies a non-sharable file.
>Execution of this command stops.
 
------ END MATRIX -----
* Note the following works:(passing from CORREL -> FACTOR).
Obviously something "special" about MATRIX (other than it has been forgotten like a bastard child)?

DATA LIST FREE/ a b c.
BEGIN DATA.
1 2 3 2 3 4 3 4 5 1 6 3 7 2 1 6 3 1 2 3 1 3 8 2
END DATA.
DATASET NAME abc.
DATASET DECLARE results .
DATASET ACTIVATE abc.

CORRELATIONS VAR a b c / MATRIX= OUT(results).
FACTOR MATRIX =IN(COR=results).


Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

Bruce Weaver
Administrator
Thanks for raising this again, David.  I agree that it is a serious problem.


David Marso wrote
The simplest most OBVIOUS code aborts with a truly inexplicable/inexcusable error!
This came up awhile ago, but no acceptable solution has been posted:
One "solution" which "works" but is unacceptable for any code running on machines where the user does not have administrative rights was posted.
Yes, running as admin 'fixes' the problem, but the solutions I and many others provide to clients
CANNOT depend or REQUIRE that the user has admin rights.  
Nor should we have to document right click>run as administrator prior to starting SPSS.
What is the point of supporting multiple datasets if they cannot be reliably used across ALL procedures.
I know another solution is to write to disk, But I really HATE writing ANYTHING to a users disk unless it is absolutely necessary.  Besides, I have important code which WORKED in ver 21 without Admin rights.
I hope this will be fixed in 22.01 or .??  How did this ever pass QA?
Any ideas of what is actually going on and the nature of the obviously incorrect error message?
----
DATA LIST LIST / a b c.
BEGIN DATA.
1 2 3
3 3 4
2 6 5
END DATA.
DATASET NAME abc.
DATASET DECLARE results .
DATASET ACTIVATE abc.
DATASET DISPLAY.

MATRIX.
GET abc / FILE * .
COMPUTE Tabc = T(abc).
SAVE Tabc / OUTFILE results .
END MATRIX.
DATASET DISPLAY.

Run MATRIX procedure:
 
>Error # 34 in column 21.  Text: results
>SPSS Statistics cannot access a file with the given file specification.  The
>file specification is either syntactically invalid, specifies an invalid
>drive, specifies a protected directory, specifies a protected file, or
>specifies a non-sharable file.
>Execution of this command stops.
 
------ END MATRIX -----
* Note the following works:(passing from CORREL -> FACTOR).
Obviously something "special" about MATRIX (other than it has been forgotten like a bastard child)?

DATA LIST FREE/ a b c.
BEGIN DATA.
1 2 3 2 3 4 3 4 5 1 6 3 7 2 1 6 3 1 2 3 1 3 8 2
END DATA.
DATASET NAME abc.
DATASET DECLARE results .
DATASET ACTIVATE abc.

CORRELATIONS VAR a b c / MATRIX= OUT(results).
FACTOR MATRIX =IN(COR=results).
--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

Marta Garcia-Granero
In reply to this post by David Marso
Hi David:

Try NOT declaring the results DATASET in advance:

DATA LIST LIST / a b c.
BEGIN DATA.
1 2 3
3 3 4
2 6 5
END DATA.
DATASET NAME abc.
DATASET DISPLAY.
MATRIX.
GET abc / FILE *.
COMPUTE Tabc = T(abc).
SAVE Tabc / OUTFILE * .
END MATRIX.
DATASET NAME results.
DATASET DISPLAY.

It looks that the MATRIX OUTFILE goes to an unnamed dataset (active),
that can be named immediately  afterwards. Unfortunately, this will not
work with MATRIX procedures that create more than one dataset during
execution.

Best regards,
Marta

(PD: Can anybody recommend a GOOOOD book on CFA, besides the one I
already have (Tabacknich&Fidell)?

El 28/10/2013 20:34, David Marso escribió:

> The simplest most OBVIOUS code aborts with a truly inexplicable/inexcusable
> error!
> This came up awhile ago, but no acceptable solution has been posted:
> One "solution" which "works" but is unacceptable for any code running on
> machines where the user does not have administrative rights was posted.
> Yes, running as admin 'fixes' the problem, but the solutions I and many
> others provide to clients
> CANNOT depend or REQUIRE that the user has admin rights.
> Nor should we have to document right click>run as administrator prior to
> starting SPSS.
> What is the point of supporting multiple datasets if they cannot be reliably
> used across ALL procedures.
> I know another solution is to write to disk, But I really HATE writing
> ANYTHING to a users disk unless it is absolutely necessary.  Besides, I have
> important code which WORKED in ver 21 without Admin rights.
> I hope this will be fixed in 22.01 or .??  How did this ever pass QA?
> Any ideas of what is actually going on and the nature of the obviously
> incorrect error message?
> ----
> DATA LIST LIST / a b c.
> BEGIN DATA.
> 1 2 3
> 3 3 4
> 2 6 5
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
> DATASET DISPLAY.
>
> MATRIX.
> GET abc / FILE * .
> COMPUTE Tabc = T(abc).
> SAVE Tabc / OUTFILE results .
> END MATRIX.
> DATASET DISPLAY.
>
> Run MATRIX procedure:
>
>> Error # 34 in column 21.  Text: results
>> SPSS Statistics cannot access a file with the given file specification.
> The
>> file specification is either syntactically invalid, specifies an invalid
>> drive, specifies a protected directory, specifies a protected file, or
>> specifies a non-sharable file.
>> Execution of this command stops.
> ------ END MATRIX -----
> * Note the following works:(passing from CORREL -> FACTOR).
> Obviously something "special" about MATRIX (other than it has been forgotten
> like a bastard child)?
>
> DATA LIST FREE/ a b c.
> BEGIN DATA.
> 1 2 3 2 3 4 3 4 5 1 6 3 7 2 1 6 3 1 2 3 1 3 8 2
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
>
> CORRELATIONS VAR a b c / MATRIX= OUT(results).
> FACTOR MATRIX =IN(COR=results).
>
>
>
>
>
>
> -----
> Please reply to the list and not to my personal email.
> Those desiring my consulting or training services please feel free to email me.
> ---
> "Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
> Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
> --
> View this message in context: http://spssx-discussion.1045642.n5.nabble.com/SERIOUS-problem-with-Multiple-datasets-in-MATRIX-procedure-tp5722765.html
> Sent from the SPSSX Discussion mailing list archive at Nabble.com.
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (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
[hidden email] (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
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

Bruce Weaver
Administrator
Hi Marta.  That is exactly what I found I had to do for something I was doing recently.  But as you say, "this will not work with MATRIX procedures that create more than one dataset during execution", which is the real problem.  

Re books on CFA, sorry, nothing comes to mind.

Bruce


Marta García-Granero-2 wrote
Hi David:

Try NOT declaring the results DATASET in advance:

DATA LIST LIST / a b c.
BEGIN DATA.
1 2 3
3 3 4
2 6 5
END DATA.
DATASET NAME abc.
DATASET DISPLAY.
MATRIX.
GET abc / FILE *.
COMPUTE Tabc = T(abc).
SAVE Tabc / OUTFILE * .
END MATRIX.
DATASET NAME results.
DATASET DISPLAY.

It looks that the MATRIX OUTFILE goes to an unnamed dataset (active),
that can be named immediately  afterwards. Unfortunately, this will not
work with MATRIX procedures that create more than one dataset during
execution.

Best regards,
Marta

(PD: Can anybody recommend a GOOOOD book on CFA, besides the one I
already have (Tabacknich&Fidell)?

El 28/10/2013 20:34, David Marso escribió:
> The simplest most OBVIOUS code aborts with a truly inexplicable/inexcusable
> error!
> This came up awhile ago, but no acceptable solution has been posted:
> One "solution" which "works" but is unacceptable for any code running on
> machines where the user does not have administrative rights was posted.
> Yes, running as admin 'fixes' the problem, but the solutions I and many
> others provide to clients
> CANNOT depend or REQUIRE that the user has admin rights.
> Nor should we have to document right click>run as administrator prior to
> starting SPSS.
> What is the point of supporting multiple datasets if they cannot be reliably
> used across ALL procedures.
> I know another solution is to write to disk, But I really HATE writing
> ANYTHING to a users disk unless it is absolutely necessary.  Besides, I have
> important code which WORKED in ver 21 without Admin rights.
> I hope this will be fixed in 22.01 or .??  How did this ever pass QA?
> Any ideas of what is actually going on and the nature of the obviously
> incorrect error message?
> ----
> DATA LIST LIST / a b c.
> BEGIN DATA.
> 1 2 3
> 3 3 4
> 2 6 5
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
> DATASET DISPLAY.
>
> MATRIX.
> GET abc / FILE * .
> COMPUTE Tabc = T(abc).
> SAVE Tabc / OUTFILE results .
> END MATRIX.
> DATASET DISPLAY.
>
> Run MATRIX procedure:
>
>> Error # 34 in column 21.  Text: results
>> SPSS Statistics cannot access a file with the given file specification.
> The
>> file specification is either syntactically invalid, specifies an invalid
>> drive, specifies a protected directory, specifies a protected file, or
>> specifies a non-sharable file.
>> Execution of this command stops.
> ------ END MATRIX -----
> * Note the following works:(passing from CORREL -> FACTOR).
> Obviously something "special" about MATRIX (other than it has been forgotten
> like a bastard child)?
>
> DATA LIST FREE/ a b c.
> BEGIN DATA.
> 1 2 3 2 3 4 3 4 5 1 6 3 7 2 1 6 3 1 2 3 1 3 8 2
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
>
> CORRELATIONS VAR a b c / MATRIX= OUT(results).
> FACTOR MATRIX =IN(COR=results).
>
>
>
>
>
>
> -----
> Please reply to the list and not to my personal email.
> Those desiring my consulting or training services please feel free to email me.
> ---
> "Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
> Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
> --
> View this message in context: http://spssx-discussion.1045642.n5.nabble.com/SERIOUS-problem-with-Multiple-datasets-in-MATRIX-procedure-tp5722765.html
> Sent from the SPSSX Discussion mailing list archive at Nabble.com.
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (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
[hidden email] (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
--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

Ware, William B
Have you seen Brown's book on CFA?  I think it's a Guilford product...

http://www.amazon.com/Confirmatory-Analysis-Research-Methodology-Sciences/dp/1593852746/ref=sr_1_1?s=books&ie=UTF8&qid=1383046133&sr=1-1&keywords=brown+confirmatory+factor+analysis

Bill

William B. Ware, Ph.D.� � � �
Professor of Learning Sciences and Psychological Studies
Adjunct Professor, School of Social Work
McMichael Term Professor of Education 2011-2013
Academy of Distinguished Teaching Scholars at UNC-Chapel Hill – Charter Member
CB #3500 - 118 Peabody Hall�
University of North Carolina at Chapel Hill
Chapel Hill, NC� � � �  27599-3500
Office:� �  (919)-962-0132�  (Sharon Powell - 309 Peabody)
Fax:� � � � � � �  (919)-962-1533
Office:�  118 Peabody Hall
EMAIL: [hidden email]



-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Bruce Weaver
Sent: Tuesday, October 29, 2013 7:21 AM
To: [hidden email]
Subject: Re: SERIOUS problem with Multiple datasets in MATRIX procedure

Hi Marta.  That is exactly what I found I had to do for something I was doing recently.  But as you say, "this will not work with MATRIX procedures that create more than one dataset during execution", which is the real problem.

Re books on CFA, sorry, nothing comes to mind.

Bruce



Marta García-Granero-2 wrote

> Hi David:
>
> Try NOT declaring the results DATASET in advance:
>
> DATA LIST LIST / a b c.
> BEGIN DATA.
> 1 2 3
> 3 3 4
> 2 6 5
> END DATA.
> DATASET NAME abc.
> DATASET DISPLAY.
> MATRIX.
> GET abc / FILE *.
> COMPUTE Tabc = T(abc).
> SAVE Tabc / OUTFILE * .
> END MATRIX.
> DATASET NAME results.
> DATASET DISPLAY.
>
> It looks that the MATRIX OUTFILE goes to an unnamed dataset (active),
> that can be named immediately  afterwards. Unfortunately, this will
> not work with MATRIX procedures that create more than one dataset
> during execution.
>
> Best regards,
> Marta
>
> (PD: Can anybody recommend a GOOOOD book on CFA, besides the one I
> already have (Tabacknich&Fidell)?
>
> El 28/10/2013 20:34, David Marso escribió:
>> The simplest most OBVIOUS code aborts with a truly
>> inexplicable/inexcusable error!
>> This came up awhile ago, but no acceptable solution has been posted:
>> One "solution" which "works" but is unacceptable for any code running
>> on machines where the user does not have administrative rights was posted.
>> Yes, running as admin 'fixes' the problem, but the solutions I and
>> many others provide to clients CANNOT depend or REQUIRE that the user
>> has admin rights.
>> Nor should we have to document right click>run as administrator prior
>> to starting SPSS.
>> What is the point of supporting multiple datasets if they cannot be
>> reliably used across ALL procedures.
>> I know another solution is to write to disk, But I really HATE
>> writing ANYTHING to a users disk unless it is absolutely necessary.
>> Besides, I have important code which WORKED in ver 21 without Admin
>> rights.
>> I hope this will be fixed in 22.01 or .??  How did this ever pass QA?
>> Any ideas of what is actually going on and the nature of the
>> obviously incorrect error message?
>> ----
>> DATA LIST LIST / a b c.
>> BEGIN DATA.
>> 1 2 3
>> 3 3 4
>> 2 6 5
>> END DATA.
>> DATASET NAME abc.
>> DATASET DECLARE results .
>> DATASET ACTIVATE abc.
>> DATASET DISPLAY.
>>
>> MATRIX.
>> GET abc / FILE * .
>> COMPUTE Tabc = T(abc).
>> SAVE Tabc / OUTFILE results .
>> END MATRIX.
>> DATASET DISPLAY.
>>
>> Run MATRIX procedure:
>>
>>> Error # 34 in column 21.  Text: results SPSS Statistics cannot
>>> access a file with the given file specification.
>> The
>>> file specification is either syntactically invalid, specifies an
>>> invalid drive, specifies a protected directory, specifies a
>>> protected file, or specifies a non-sharable file.
>>> Execution of this command stops.
>> ------ END MATRIX -----
>> * Note the following works:(passing from CORREL -> FACTOR).
>> Obviously something "special" about MATRIX (other than it has been
>> forgotten like a bastard child)?
>>
>> DATA LIST FREE/ a b c.
>> BEGIN DATA.
>> 1 2 3 2 3 4 3 4 5 1 6 3 7 2 1 6 3 1 2 3 1 3 8 2 END DATA.
>> DATASET NAME abc.
>> DATASET DECLARE results .
>> DATASET ACTIVATE abc.
>>
>> CORRELATIONS VAR a b c / MATRIX= OUT(results).
>> FACTOR MATRIX =IN(COR=results).
>>
>>
>>
>>
>>
>>
>> -----
>> Please reply to the list and not to my personal email.
>> Those desiring my consulting or training services please feel free to
>> email me.
>> ---
>> "Nolite dare sanctum canibus neque mittatis margaritas vestras ante
>> porcos ne forte conculcent eas pedibus suis."
>> Cum es damnatorum possederunt porcos iens ut salire off sanguinum
>> cliff in abyssum?"
>> --
>> View this message in context:
>> http://spssx-discussion.1045642.n5.nabble.com/SERIOUS-problem-with-Mu
>> ltiple-datasets-in-MATRIX-procedure-tp5722765.html
>> Sent from the SPSSX Discussion mailing list archive at Nabble.com.
>>
>> =====================
>> To manage your subscription to SPSSX-L, send a message to
>>

> LISTSERV@.UGA

>  (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@.UGA

>  (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





-----
--
Bruce Weaver
[hidden email]
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.

--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/SERIOUS-problem-with-Multiple-datasets-in-MATRIX-procedure-tp5722765p5722774.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to [hidden email] (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
[hidden email] (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
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

David Marso
Administrator
In reply to this post by Marta Garcia-Granero
Hi Marta,

"Unfortunately, this will not work with MATRIX procedures that create more than one dataset during
execution. "

It also has the terrible side effect of overwriting the active data file.
In fact, the code I have been creating does create multiple (sometimes several) output data sets ;-(

It worked perfectly under development in ver 21 and now it has gone into the crapper for 22 unless I run as admin.
IBM? PLEASE FIX THIS! IT WORKED, YOU BROKE IT.  Probably some noob-dev touching delicate code on a Friday after 5PM.
-----
(PD: Can anybody recommend a GOOOOD book on CFA, besides the one I
already have (Tabacknich&Fidell)?

Touchy subject!  
Just lurk in the SEMNET group for a week or a month (be sure to don some mud flaps).
https://listserv.ua.edu/cgi-bin/wa?A0=SEMNET
after awhile and you will note surprisingly little consensus about SEM methodology.
Sometimes (unfortunately) it gets rather personal and ugly.  
Still enough signal to noise ratio to keep it in my digest subscriptions, but seriously people, take your personal BS into IM or email.  I have been tempted to start diagramming the various factions with little balance theory triangles.  It would be amusing to attend an SEM conference to see if a gang fight breaks out ;-).

Any choice of book is going to be dictated by what you want to learn and the level/depth you wish to explore.  At what level are you trying to approach it?  Back when I was avidly researching SEM for grad school projects (circa 1985-87), some of the most informative, seminal work was being published in journals.  Most of it unlikely to hit any book until 3-5 years later.  Back then it was a fairly small field.  These days everybody and their grandmother is tooting some book about SEM.  Back then we had the LISREL manual, LISREL on the mainframe and whatever journal articles we could find.
If I were buying a book for myself, I would avoid any software-centric books  
SEM for Amos users, SEM for MPlus users etc...and go for the meat/theory.  
I personally would stray from tomes which belabor the intricacies of parameter estimation and algorithmic snooze.  I am not intending to write an SEM program, so TMI!  
I would also avoid the all too common SEM for noobs type books as they pitch to the lowest common denominator, and are unlikely to provide any intellectual nourishment to people experienced in MVA.

With that said:  If I were to buy only one book on SEM I might consider:
http://www.amazon.com/Handbook-Structural-Equation-Modeling-Hoyle/dp/1606230778/ref=sr_1_1?ie=UTF8&qid=1383047132&sr=8-1&keywords=Handbook+of+Structural+Equations+modeling
I don't own it and haven't read it, but the list of contributors is pretty impressive.  After that, you will have plenty of references to continue further reading.

BTW:  I took courses from both Barb Tabachnick and Linda Fidel back in the day at CSUN.



Marta García-Granero-2 wrote
Hi David:

Try NOT declaring the results DATASET in advance:

DATA LIST LIST / a b c.
BEGIN DATA.
1 2 3
3 3 4
2 6 5
END DATA.
DATASET NAME abc.
DATASET DISPLAY.
MATRIX.
GET abc / FILE *.
COMPUTE Tabc = T(abc).
SAVE Tabc / OUTFILE * .
END MATRIX.
DATASET NAME results.
DATASET DISPLAY.

It looks that the MATRIX OUTFILE goes to an unnamed dataset (active),
that can be named immediately  afterwards. Unfortunately, this will not
work with MATRIX procedures that create more than one dataset during
execution.

Best regards,
Marta

(PD: Can anybody recommend a GOOOOD book on CFA, besides the one I
already have (Tabacknich&Fidell)?

El 28/10/2013 20:34, David Marso escribió:
> The simplest most OBVIOUS code aborts with a truly inexplicable/inexcusable
> error!
> This came up awhile ago, but no acceptable solution has been posted:
> One "solution" which "works" but is unacceptable for any code running on
> machines where the user does not have administrative rights was posted.
> Yes, running as admin 'fixes' the problem, but the solutions I and many
> others provide to clients
> CANNOT depend or REQUIRE that the user has admin rights.
> Nor should we have to document right click>run as administrator prior to
> starting SPSS.
> What is the point of supporting multiple datasets if they cannot be reliably
> used across ALL procedures.
> I know another solution is to write to disk, But I really HATE writing
> ANYTHING to a users disk unless it is absolutely necessary.  Besides, I have
> important code which WORKED in ver 21 without Admin rights.
> I hope this will be fixed in 22.01 or .??  How did this ever pass QA?
> Any ideas of what is actually going on and the nature of the obviously
> incorrect error message?
> ----
> DATA LIST LIST / a b c.
> BEGIN DATA.
> 1 2 3
> 3 3 4
> 2 6 5
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
> DATASET DISPLAY.
>
> MATRIX.
> GET abc / FILE * .
> COMPUTE Tabc = T(abc).
> SAVE Tabc / OUTFILE results .
> END MATRIX.
> DATASET DISPLAY.
>
> Run MATRIX procedure:
>
>> Error # 34 in column 21.  Text: results
>> SPSS Statistics cannot access a file with the given file specification.
> The
>> file specification is either syntactically invalid, specifies an invalid
>> drive, specifies a protected directory, specifies a protected file, or
>> specifies a non-sharable file.
>> Execution of this command stops.
> ------ END MATRIX -----
> * Note the following works:(passing from CORREL -> FACTOR).
> Obviously something "special" about MATRIX (other than it has been forgotten
> like a bastard child)?
>
> DATA LIST FREE/ a b c.
> BEGIN DATA.
> 1 2 3 2 3 4 3 4 5 1 6 3 7 2 1 6 3 1 2 3 1 3 8 2
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
>
> CORRELATIONS VAR a b c / MATRIX= OUT(results).
> FACTOR MATRIX =IN(COR=results).
>
>
>
>
>
>
> -----
> Please reply to the list and not to my personal email.
> Those desiring my consulting or training services please feel free to email me.
> ---
> "Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
> Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
> --
> View this message in context: http://spssx-discussion.1045642.n5.nabble.com/SERIOUS-problem-with-Multiple-datasets-in-MATRIX-procedure-tp5722765.html
> Sent from the SPSSX Discussion mailing list archive at Nabble.com.
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (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
[hidden email] (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
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

David Marso
Administrator
In reply to this post by Bruce Weaver
In my (NSH) opinion it is completely unacceptable! BUG BUG BUG BUG !!!!!!
I can't wait to get the emails from people using my program (which WORKED PREVIOUSLY).
YES I AM SHOUTING! CAN ANYBODY AT IBM HEAR ME? DO YOU GIVE A HOOT?
Screw it, I'll just dump everything to disk.  What is the point of multiple data sets?
AGGREGATE?  Beyond that please enlighten me!~  
Seems like a lot of dev-effort resulting in a half assed implementation!

Bruce Weaver wrote
Hi Marta.  That is exactly what I found I had to do for something I was doing recently.  But as you say, "this will not work with MATRIX procedures that create more than one dataset during execution", which is the real problem.  

Re books on CFA, sorry, nothing comes to mind.

Bruce


Marta García-Granero-2 wrote
Hi David:

Try NOT declaring the results DATASET in advance:

DATA LIST LIST / a b c.
BEGIN DATA.
1 2 3
3 3 4
2 6 5
END DATA.
DATASET NAME abc.
DATASET DISPLAY.
MATRIX.
GET abc / FILE *.
COMPUTE Tabc = T(abc).
SAVE Tabc / OUTFILE * .
END MATRIX.
DATASET NAME results.
DATASET DISPLAY.

It looks that the MATRIX OUTFILE goes to an unnamed dataset (active),
that can be named immediately  afterwards. Unfortunately, this will not
work with MATRIX procedures that create more than one dataset during
execution.

Best regards,
Marta

(PD: Can anybody recommend a GOOOOD book on CFA, besides the one I
already have (Tabacknich&Fidell)?

El 28/10/2013 20:34, David Marso escribió:
> The simplest most OBVIOUS code aborts with a truly inexplicable/inexcusable
> error!
> This came up awhile ago, but no acceptable solution has been posted:
> One "solution" which "works" but is unacceptable for any code running on
> machines where the user does not have administrative rights was posted.
> Yes, running as admin 'fixes' the problem, but the solutions I and many
> others provide to clients
> CANNOT depend or REQUIRE that the user has admin rights.
> Nor should we have to document right click>run as administrator prior to
> starting SPSS.
> What is the point of supporting multiple datasets if they cannot be reliably
> used across ALL procedures.
> I know another solution is to write to disk, But I really HATE writing
> ANYTHING to a users disk unless it is absolutely necessary.  Besides, I have
> important code which WORKED in ver 21 without Admin rights.
> I hope this will be fixed in 22.01 or .??  How did this ever pass QA?
> Any ideas of what is actually going on and the nature of the obviously
> incorrect error message?
> ----
> DATA LIST LIST / a b c.
> BEGIN DATA.
> 1 2 3
> 3 3 4
> 2 6 5
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
> DATASET DISPLAY.
>
> MATRIX.
> GET abc / FILE * .
> COMPUTE Tabc = T(abc).
> SAVE Tabc / OUTFILE results .
> END MATRIX.
> DATASET DISPLAY.
>
> Run MATRIX procedure:
>
>> Error # 34 in column 21.  Text: results
>> SPSS Statistics cannot access a file with the given file specification.
> The
>> file specification is either syntactically invalid, specifies an invalid
>> drive, specifies a protected directory, specifies a protected file, or
>> specifies a non-sharable file.
>> Execution of this command stops.
> ------ END MATRIX -----
> * Note the following works:(passing from CORREL -> FACTOR).
> Obviously something "special" about MATRIX (other than it has been forgotten
> like a bastard child)?
>
> DATA LIST FREE/ a b c.
> BEGIN DATA.
> 1 2 3 2 3 4 3 4 5 1 6 3 7 2 1 6 3 1 2 3 1 3 8 2
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
>
> CORRELATIONS VAR a b c / MATRIX= OUT(results).
> FACTOR MATRIX =IN(COR=results).
>
>
>
>
>
>
> -----
> Please reply to the list and not to my personal email.
> Those desiring my consulting or training services please feel free to email me.
> ---
> "Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
> Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
> --
> View this message in context: http://spssx-discussion.1045642.n5.nabble.com/SERIOUS-problem-with-Multiple-datasets-in-MATRIX-procedure-tp5722765.html
> Sent from the SPSSX Discussion mailing list archive at Nabble.com.
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (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
[hidden email] (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
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

David Dwyer
In reply to this post by David Marso
 Hello David,
 
I have tested your code in Statistics 22.0 GA on Windows 7 and
Windows 8.1 Preview.  I can confirm your diagnosis that it appears
to be permissions related:

 1) Your code works fine when Statistics has been launched with "Run
as Administrator".

 2) Your code works fine if the local "Users" security group has been
given "Modify" permission to the Statistics installation directory.

 
I've logged a defect on this issue with IBM SPSS Statistics
Development and hope to have a resolution in a future fixpack.

 
I apologize for the inconvenience.

 

David J. Dwyer
IBM SPSS Solution Consultant
Business Analytics

Tel: 312-477-4542
E-mail: [hidden email]

 

 
> From: David Marso <[hidden email]

> To: [hidden email],
> Date: 10/28/2013 02:35 PM
> Subject: SERIOUS problem with Multiple datasets in MATRIX procedure
> Sent by: "SPSSX(r) Discussion" <[hidden email]>
>
> The simplest most OBVIOUS code aborts with a truly inexplicable/inexcusable

> error!
> This came up awhile ago, but no acceptable solution has been posted:
> One "solution" which "works" but is unacceptable for any code running on
> machines where the user does not have administrative rights was posted.
> Yes, running as admin 'fixes' the problem, but the solutions I and many
> others provide to clients
> CANNOT depend or REQUIRE that the user has admin rights.
> Nor should we have to document right click>run as administrator prior to
> starting SPSS.
> What is the point of supporting multiple datasets if they cannot be reliably
> used across ALL procedures.
> I know another solution is to write to disk, But I really HATE writing
> ANYTHING to a users disk unless it is absolutely necessary.  Besides, I have
> important code which WORKED in ver 21 without Admin rights.
> I hope this will be fixed in 22.01 or .??  How did this ever pass QA?
> Any ideas of what is actually going on and the nature of the obviously
> incorrect error message?
> ----
> DATA LIST LIST / a b c.
> BEGIN DATA.
> 1 2 3
> 3 3 4
> 2 6 5
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
> DATASET DISPLAY.
>
> MATRIX.

> GET abc / FILE * .
> COMPUTE Tabc = T(abc).
> SAVE Tabc / OUTFILE results .
> END MATRIX.
> DATASET DISPLAY.
>
> Run MATRIX procedure:

>
> >Error # 34 in column 21.  Text: results

> >SPSS Statistics cannot access a file with the given file specification.
> The
> >file specification is either syntactically invalid, specifies an invalid
> >drive, specifies a protected directory, specifies a protected file, or
> >specifies a non-sharable file.
> >Execution of this command stops.
>
> ------ END MATRIX -----

> * Note the following works:(passing from CORREL -> FACTOR).
> Obviously something "special" about MATRIX (other than it has been forgotten
> like a bastard child)?
>
> DATA LIST FREE/ a b c.

> BEGIN DATA.
> 1 2 3 2 3 4 3 4 5 1 6 3 7 2 1 6 3 1 2 3 1 3 8 2
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
>
> CORRELATIONS VAR a b c / MATRIX= OUT(results).

> FACTOR MATRIX =IN(COR=results).
>
> -----

> Please reply to the list and not to my personal email.
> Those desiring my consulting or training services please feel free
> to email me.

> ---
> "Nolite dare sanctum canibus neque mittatis margaritas vestras ante
> porcos ne forte conculcent eas pedibus suis."

> Cum es damnatorum possederunt porcos iens ut salire off sanguinum
> cliff in abyssum?"

> --
> View this message in context: http://spssx-discussion.
> 1045642.n5.nabble.com/SERIOUS-problem-with-Multiple-datasets-in-
> MATRIX-procedure-tp5722765.html

> Sent from the SPSSX Discussion mailing list archive at Nabble.com.
>
> =====================

> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (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
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

David Marso
Administrator
Thanks David,
You know that Win 8.1 is out now.  
I have it running and quite the improvement to S=Win8.0
While you are at it do you think you could see if it would be too
dreadful to fix MATRIX  so it can read from datasets rather than
just the active file and disk.
That might be considered a feature request, but symmetry is one of the
fine beautiful things in nature.  
Not having this makes the general idea of allowing multiple datasets rather useless
for some of us.  Except perhaps for AGGREGATE.
The plan here is to use AGGREGATE to build out various summary statistics based on different break variables, pushing them into datasets.  Then use MATRIX to further manipulate them and make magical things happen.  I know I COULD stuff them onto disk, but I DON'T WANT to and shouldn't HAVE TO.
As stated previously, I don't like to touch the user's disk unless I have to.  Also if I do need to write to disk I need to write clean up code and we all love those stupid file contention issues SPSS is so happy to provide ;-)
David,
PS PM me .
---
David Dwyer wrote
Hello David,

 I have tested your code in Statistics 22.0 GA on Windows 7 and
 Windows 8.1 Preview.  I can confirm your diagnosis that it appears
 to be permissions related:
 1) Your code works fine when Statistics has been launched with "Run
 as Administrator".
 2) Your code works fine if the local "Users" security group has been
 given "Modify" permission to the Statistics installation directory.

 I've logged a defect on this issue with IBM SPSS Statistics
 Development and hope to have a resolution in a future fixpack.

 I apologize for the inconvenience.


 David J. Dwyer
 IBM SPSS Solution Consultant
 Business Analytics

 Tel: 312-477-4542
 E-mail: [hidden email]



> From: David Marso <[hidden email]
> To: [hidden email],
> Date: 10/28/2013 02:35 PM
> Subject: SERIOUS problem with Multiple datasets in MATRIX procedure
> Sent by: "SPSSX(r) Discussion" <[hidden email]>
>
> The simplest most OBVIOUS code aborts with a truly
inexplicable/inexcusable
> error!
> This came up awhile ago, but no acceptable solution has been posted:
> One "solution" which "works" but is unacceptable for any code running on
> machines where the user does not have administrative rights was posted.
> Yes, running as admin 'fixes' the problem, but the solutions I and many
> others provide to clients
> CANNOT depend or REQUIRE that the user has admin rights.
> Nor should we have to document right click>run as administrator prior to
> starting SPSS.
> What is the point of supporting multiple datasets if they cannot be
reliably
> used across ALL procedures.
> I know another solution is to write to disk, But I really HATE writing
> ANYTHING to a users disk unless it is absolutely necessary.  Besides, I
have
> important code which WORKED in ver 21 without Admin rights.
> I hope this will be fixed in 22.01 or .??  How did this ever pass QA?
> Any ideas of what is actually going on and the nature of the obviously
> incorrect error message?
> ----
> DATA LIST LIST / a b c.
> BEGIN DATA.
> 1 2 3
> 3 3 4
> 2 6 5
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
> DATASET DISPLAY.
>
> MATRIX.
> GET abc / FILE * .
> COMPUTE Tabc = T(abc).
> SAVE Tabc / OUTFILE results .
> END MATRIX.
> DATASET DISPLAY.
>
> Run MATRIX procedure:
>
> >Error # 34 in column 21.  Text: results
> >SPSS Statistics cannot access a file with the given file specification.
> The
> >file specification is either syntactically invalid, specifies an
invalid
> >drive, specifies a protected directory, specifies a protected file, or
> >specifies a non-sharable file.
> >Execution of this command stops.
>
> ------ END MATRIX -----
> * Note the following works:(passing from CORREL -> FACTOR).
> Obviously something "special" about MATRIX (other than it has been
forgotten
> like a bastard child)?
>
> DATA LIST FREE/ a b c.
> BEGIN DATA.
> 1 2 3 2 3 4 3 4 5 1 6 3 7 2 1 6 3 1 2 3 1 3 8 2
> END DATA.
> DATASET NAME abc.
> DATASET DECLARE results .
> DATASET ACTIVATE abc.
>
> CORRELATIONS VAR a b c / MATRIX= OUT(results).
> FACTOR MATRIX =IN(COR=results).
>
> -----
> Please reply to the list and not to my personal email.
> Those desiring my consulting or training services please feel free
> to email me.
> ---
> "Nolite dare sanctum canibus neque mittatis margaritas vestras ante
> porcos ne forte conculcent eas pedibus suis."
> Cum es damnatorum possederunt porcos iens ut salire off sanguinum
> cliff in abyssum?"
> --
> View this message in context: http://spssx-discussion.
> 1045642.n5.nabble.com/SERIOUS-problem-with-Multiple-datasets-in-
> MATRIX-procedure-tp5722765.html
> Sent from the SPSSX Discussion mailing list archive at Nabble.com.
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (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
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

PRogman
The problem is surely related to the initialization of the work folder. The comments in this thread  point out that 1) Matrix Datasets don't work as expected, _unless_ run in administrator mode 2) if R/W permissions are given to the SPSS installation folder it works.
Most Windows users have the User Account Control (UAC) activated (requiring manual confirmation when a program wants to access the installation folders). UAC was introduced in Windows Vista 2007 and the folder Program Files and its subfolders were better protected.
I noticed that if the default working directory (folder) is changed to a folder with R/W permission MATRIX works as expected!
SPSS defaults to the installation folder ...\ templates when handling chart templates -- most annoying as it is protected, and it does not retain the working folder when another is used. Same with scripts. It may be similar with TableLooks which I have not used. The panel 'Edit|Options|File Locations' do need more specifications in my opinion.
UAC was introduced in 2007... it should have been fixed long ago!

/PR

*-----------------------------------------------------*.
NEW FILE.
DATA LIST LIST /
a b c.
BEGIN DATA.
1 2 3
3 3 4
2 6 5
END DATA.
DATASET NAME abc.

* Windows *.
FILE HANDLE fpFolder_With_ReadWriteRights /NAME='%userprofile%\Desktop' .

* The following command will have exit code 1 and issue a warning if no matching files are found *.
* Note the initial folder name: the SPSS installation folder (as told in TFM under 'HOST': *.
* 'The HOST command starts in the current working directory.                      *.
*  By default, the initial working directory is the installation directory.'      *.

HOST COMMAND ='Dir /AD /N *.$$$' .

* Save current default folder, and change to one with R/W permission *.
PRESERVE.
CD fpFolder_With_ReadWriteRights .

* Datasets to matrix *.
DATASET DECLARE results1 .
DATASET DECLARE results2 .
DATASET DECLARE results3 .
DATASET DECLARE results4 .

DATASET ACTIVATE abc.
DATASET DISPLAY.

MATRIX.
  GET abc / FILE * .
  COMPUTE Tabc = T(abc).
  SAVE Tabc / OUTFILE results1 .
  SAVE Tabc / OUTFILE results2 .
END MATRIX.

DATASET DISPLAY.

* Restore the initial default folder: probably the SPSS installation folder *.
RESTORE.

* Matrix blues *.
MATRIX.
  GET abc / FILE * .
  COMPUTE Tabc = T(abc).
  SAVE Tabc / OUTFILE results3 .
  SAVE Tabc / OUTFILE results4 .
END MATRIX.

DATASET DISPLAY.

 * DATASET CLOSE results1 .
 * DATASET CLOSE results2 .
 * DATASET CLOSE results3 .
 * DATASET CLOSE results4 .
*-----------------------------------------------------*.
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

Bruce Weaver
Administrator
Very nice, PR.  Thanks for posting this!  


PRogman wrote
The problem is surely related to the initialization of the work folder. The comments in this thread  point out that 1) Matrix Datasets don't work as expected, _unless_ run in administrator mode 2) if R/W permissions are given to the SPSS installation folder it works.
Most Windows users have the User Account Control (UAC) activated (requiring manual confirmation when a program wants to access the installation folders). UAC was introduced in Windows Vista 2007 and the folder Program Files and its subfolders were better protected.
I noticed that if the default working directory (folder) is changed to a folder with R/W permission MATRIX works as expected!
SPSS defaults to the installation folder ...\ templates when handling chart templates -- most annoying as it is protected, and it does not retain the working folder when another is used. Same with scripts. It may be similar with TableLooks which I have not used. The panel 'Edit|Options|File Locations' do need more specifications in my opinion.
UAC was introduced in 2007... it should have been fixed long ago!

/PR

*-----------------------------------------------------*.
NEW FILE.
DATA LIST LIST /
a b c.
BEGIN DATA.
1 2 3
3 3 4
2 6 5
END DATA.
DATASET NAME abc.

* Windows *.
FILE HANDLE fpFolder_With_ReadWriteRights /NAME='%userprofile%\Desktop' .

* The following command will have exit code 1 and issue a warning if no matching files are found *.
* Note the initial folder name: the SPSS installation folder (as told in TFM under 'HOST': *.
* 'The HOST command starts in the current working directory.                      *.
*  By default, the initial working directory is the installation directory.'      *.

HOST COMMAND ='Dir /AD /N *.$$$' .

* Save current default folder, and change to one with R/W permission *.
PRESERVE.
CD fpFolder_With_ReadWriteRights .

* Datasets to matrix *.
DATASET DECLARE results1 .
DATASET DECLARE results2 .
DATASET DECLARE results3 .
DATASET DECLARE results4 .

DATASET ACTIVATE abc.
DATASET DISPLAY.

MATRIX.
  GET abc / FILE * .
  COMPUTE Tabc = T(abc).
  SAVE Tabc / OUTFILE results1 .
  SAVE Tabc / OUTFILE results2 .
END MATRIX.

DATASET DISPLAY.

* Restore the initial default folder: probably the SPSS installation folder *.
RESTORE.

* Matrix blues *.
MATRIX.
  GET abc / FILE * .
  COMPUTE Tabc = T(abc).
  SAVE Tabc / OUTFILE results3 .
  SAVE Tabc / OUTFILE results4 .
END MATRIX.

DATASET DISPLAY.

 * DATASET CLOSE results1 .
 * DATASET CLOSE results2 .
 * DATASET CLOSE results3 .
 * DATASET CLOSE results4 .
*-----------------------------------------------------*.
--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

Rick Oliver-3
Edit>Options>File Locations has no effect on command syntax. It controls the default locations for file chooser dialogs. The generated/pasted syntax from those dialogs contains the full path specification.

It's interesting that using the CD command to change the working directory would have any effect on datasets used with Matrix, since theoretically datasets in the current session should be stored in whatever is defined as the temp directory, which is not the same as the working directory.

There are a number of known issues when using datasets with Matrix, but until now I thought the only solution was "don't use datasets with Matrix".

Rick Oliver
Senior Information Developer
IBM Business Analytics (SPSS)
E-mail: [hidden email]




From:        Bruce Weaver <[hidden email]>
To:        [hidden email],
Date:        04/30/2014 03:32 PM
Subject:        Re: SERIOUS problem with Multiple datasets in MATRIX procedure
Sent by:        "SPSSX(r) Discussion" <[hidden email]>




Very nice, PR.  Thanks for posting this!



PRogman wrote
> The problem is surely related to the initialization of the work folder.
> The comments in this thread  point out that 1) Matrix Datasets don't work
> as expected, _unless_ run in administrator mode 2) if R/W permissions are
> given to the SPSS installation folder it works.
> Most Windows users have the User Account Control (UAC) activated
> (requiring manual confirmation when a program wants to access the
> installation folders). UAC was introduced in Windows Vista 2007 and the
> folder Program Files and its subfolders were better protected.
> I noticed that if the default working directory (folder) is changed to a
> folder with R/W permission MATRIX works as expected!
> SPSS defaults to the installation folder ...\ templates when handling
> chart templates -- most annoying as it is protected, and it does not
> retain the working folder when another is used. Same with scripts. It may
> be similar with TableLooks which I have not used. The panel
> 'Edit|Options|File Locations' do need more specifications in my opinion.
> UAC was introduced in 2007... it should have been fixed long ago!
>
> /PR
>
> *-----------------------------------------------------*.
> NEW FILE.
> DATA LIST LIST /
> a b c.
> BEGIN DATA.
> 1 2 3
> 3 3 4
> 2 6 5
> END DATA.
> DATASET NAME abc.
>
> * Windows *.
> FILE HANDLE fpFolder_With_ReadWriteRights /NAME='%userprofile%\Desktop' .
>
> * The following command will have exit code 1 and issue a warning if no
> matching files are found *.
> * Note the initial folder name: the SPSS installation folder (as told in
> TFM under 'HOST': *.
> * 'The HOST command starts in the current working directory.
> *.
> *  By default, the initial working directory is the installation
> directory.'      *.
>
> HOST COMMAND ='Dir /AD /N *.$$$' .
>
> * Save current default folder, and change to one with R/W permission *.
> PRESERVE.
> CD fpFolder_With_ReadWriteRights .
>
> * Datasets to matrix *.
> DATASET DECLARE results1 .
> DATASET DECLARE results2 .
> DATASET DECLARE results3 .
> DATASET DECLARE results4 .
>
> DATASET ACTIVATE abc.
> DATASET DISPLAY.
>
> MATRIX.
>   GET abc / FILE * .
>   COMPUTE Tabc = T(abc).
>   SAVE Tabc / OUTFILE results1 .
>   SAVE Tabc / OUTFILE results2 .
> END MATRIX.
>
> DATASET DISPLAY.
>
> * Restore the initial default folder: probably the SPSS installation
> folder *.
> RESTORE.
>
> * Matrix blues *.
> MATRIX.
>   GET abc / FILE * .
>   COMPUTE Tabc = T(abc).
>   SAVE Tabc / OUTFILE results3 .
>   SAVE Tabc / OUTFILE results4 .
> END MATRIX.
>
> DATASET DISPLAY.
>
>  * DATASET CLOSE results1 .
>  * DATASET CLOSE results2 .
>  * DATASET CLOSE results3 .
>  * DATASET CLOSE results4 .
> *-----------------------------------------------------*.





-----
--
Bruce Weaver
[hidden email]
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.

--
View this message in context:
http://spssx-discussion.1045642.n5.nabble.com/SERIOUS-problem-with-Multiple-datasets-in-MATRIX-procedure-tp5722765p5725745.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (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


Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

David Marso
Administrator
"There are a number of known issues when using datasets with Matrix, but until now I thought the only solution was "don't use datasets with Matrix". "
Can you provide us with a LIST of these known issues so people such as Kirill, Brian and myself don't waste our time trying to find workarounds?
I don't believe 'don't use datasets with MATRIX' is an acceptable 'solution'!!!!!

Rick Oliver wrote
Edit>Options>File Locations has no effect on command syntax. It controls
the default locations for file chooser dialogs. The generated/pasted
syntax from those dialogs contains the full path specification.

It's interesting that using the CD command to change the working directory
would have any effect on datasets used with Matrix, since theoretically
datasets in the current session should be stored in whatever is defined as
the temp directory, which is not the same as the working directory.

There are a number of known issues when using datasets with Matrix, but
until now I thought the only solution was "don't use datasets with
Matrix".

Rick Oliver
Senior Information Developer
IBM Business Analytics (SPSS)
E-mail: [hidden email]



From:   Bruce Weaver <[hidden email]>
To:     [hidden email],
Date:   04/30/2014 03:32 PM
Subject:        Re: SERIOUS problem with Multiple datasets in MATRIX
procedure
Sent by:        "SPSSX(r) Discussion" <[hidden email]>



Very nice, PR.  Thanks for posting this!



PRogman wrote
> The problem is surely related to the initialization of the work folder.
> The comments in this thread  point out that 1) Matrix Datasets don't
work
> as expected, _unless_ run in administrator mode 2) if R/W permissions
are
> given to the SPSS installation folder it works.
> Most Windows users have the User Account Control (UAC) activated
> (requiring manual confirmation when a program wants to access the
> installation folders). UAC was introduced in Windows Vista 2007 and the
> folder Program Files and its subfolders were better protected.
> I noticed that if the default working directory (folder) is changed to a
> folder with R/W permission MATRIX works as expected!
> SPSS defaults to the installation folder ...\ templates when handling
> chart templates -- most annoying as it is protected, and it does not
> retain the working folder when another is used. Same with scripts. It
may
> be similar with TableLooks which I have not used. The panel
> 'Edit|Options|File Locations' do need more specifications in my opinion.
> UAC was introduced in 2007... it should have been fixed long ago!
>
> /PR
>
> *-----------------------------------------------------*.
> NEW FILE.
> DATA LIST LIST /
> a b c.
> BEGIN DATA.
> 1 2 3
> 3 3 4
> 2 6 5
> END DATA.
> DATASET NAME abc.
>
> * Windows *.
> FILE HANDLE fpFolder_With_ReadWriteRights /NAME='%userprofile%\Desktop'
.
>
> * The following command will have exit code 1 and issue a warning if no
> matching files are found *.
> * Note the initial folder name: the SPSS installation folder (as told in
> TFM under 'HOST': *.
> * 'The HOST command starts in the current working directory.
> *.
> *  By default, the initial working directory is the installation
> directory.'      *.
>
> HOST COMMAND ='Dir /AD /N *.$$$' .
>
> * Save current default folder, and change to one with R/W permission *.
> PRESERVE.
> CD fpFolder_With_ReadWriteRights .
>
> * Datasets to matrix *.
> DATASET DECLARE results1 .
> DATASET DECLARE results2 .
> DATASET DECLARE results3 .
> DATASET DECLARE results4 .
>
> DATASET ACTIVATE abc.
> DATASET DISPLAY.
>
> MATRIX.
>   GET abc / FILE * .
>   COMPUTE Tabc = T(abc).
>   SAVE Tabc / OUTFILE results1 .
>   SAVE Tabc / OUTFILE results2 .
> END MATRIX.
>
> DATASET DISPLAY.
>
> * Restore the initial default folder: probably the SPSS installation
> folder *.
> RESTORE.
>
> * Matrix blues *.
> MATRIX.
>   GET abc / FILE * .
>   COMPUTE Tabc = T(abc).
>   SAVE Tabc / OUTFILE results3 .
>   SAVE Tabc / OUTFILE results4 .
> END MATRIX.
>
> DATASET DISPLAY.
>
>  * DATASET CLOSE results1 .
>  * DATASET CLOSE results2 .
>  * DATASET CLOSE results3 .
>  * DATASET CLOSE results4 .
> *-----------------------------------------------------*.





-----
--
Bruce Weaver
[hidden email]
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.

--
View this message in context:
http://spssx-discussion.1045642.n5.nabble.com/SERIOUS-problem-with-Multiple-datasets-in-MATRIX-procedure-tp5722765p5725745.html

Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (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
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
Reply | Threaded
Open this post in threaded view
|

Re: SERIOUS problem with Multiple datasets in MATRIX procedure

Rick Oliver-3
At this point, the only known workaround is to use files instead of datasets -- but this new information about changing the working data file may change that.

Rick Oliver
Senior Information Developer
IBM Business Analytics (SPSS)
E-mail: [hidden email]




From:        David Marso <[hidden email]>
To:        [hidden email],
Date:        04/30/2014 04:23 PM
Subject:        Re: SERIOUS problem with Multiple datasets in MATRIX procedure
Sent by:        "SPSSX(r) Discussion" <[hidden email]>




"There are a number of known issues when using datasets with Matrix, but
until now I thought the only solution was "don't use datasets with Matrix".
"
Can you provide us with a LIST of these known issues so people such as
Kirill, Brian and myself don't waste our time trying to find workarounds?
I don't believe 'don't use datasets with MATRIX' is an acceptable
'solution'!!!!!


Rick Oliver wrote
> Edit>Options>File Locations has no effect on command syntax. It controls
> the default locations for file chooser dialogs. The generated/pasted
> syntax from those dialogs contains the full path specification.
>
> It's interesting that using the CD command to change the working directory
> would have any effect on datasets used with Matrix, since theoretically
> datasets in the current session should be stored in whatever is defined as
> the temp directory, which is not the same as the working directory.
>
> There are a number of known issues when using datasets with Matrix, but
> until now I thought the only solution was "don't use datasets with
> Matrix".
>
> Rick Oliver
> Senior Information Developer
> IBM Business Analytics (SPSS)
> E-mail:

> oliverr@.ibm

>
>
>
> From:   Bruce Weaver &lt;

> bruce.weaver@

> &gt;
> To:

> SPSSX-L@.uga

> ,
> Date:   04/30/2014 03:32 PM
> Subject:        Re: SERIOUS problem with Multiple datasets in MATRIX
> procedure
> Sent by:        "SPSSX(r) Discussion" &lt;

> SPSSX-L@.uga

> &gt;
>
>
>
> Very nice, PR.  Thanks for posting this!
>
>
>
> PRogman wrote
>> The problem is surely related to the initialization of the work folder.
>> The comments in this thread  point out that 1) Matrix Datasets don't
> work
>> as expected, _unless_ run in administrator mode 2) if R/W permissions
> are
>> given to the SPSS installation folder it works.
>> Most Windows users have the User Account Control (UAC) activated
>> (requiring manual confirmation when a program wants to access the
>> installation folders). UAC was introduced in Windows Vista 2007 and the
>> folder Program Files and its subfolders were better protected.
>> I noticed that if the default working directory (folder) is changed to a
>> folder with R/W permission MATRIX works as expected!
>> SPSS defaults to the installation folder ...\ templates when handling
>> chart templates -- most annoying as it is protected, and it does not
>> retain the working folder when another is used. Same with scripts. It
> may
>> be similar with TableLooks which I have not used. The panel
>> 'Edit|Options|File Locations' do need more specifications in my opinion.
>> UAC was introduced in 2007... it should have been fixed long ago!
>>
>> /PR
>>
>> *-----------------------------------------------------*.
>> NEW FILE.
>> DATA LIST LIST /
>> a b c.
>> BEGIN DATA.
>> 1 2 3
>> 3 3 4
>> 2 6 5
>> END DATA.
>> DATASET NAME abc.
>>
>> * Windows *.
>> FILE HANDLE fpFolder_With_ReadWriteRights /NAME='%userprofile%\Desktop'
> .
>>
>> * The following command will have exit code 1 and issue a warning if no
>> matching files are found *.
>> * Note the initial folder name: the SPSS installation folder (as told in
>> TFM under 'HOST': *.
>> * 'The HOST command starts in the current working directory.
>> *.
>> *  By default, the initial working directory is the installation
>> directory.'      *.
>>
>> HOST COMMAND ='Dir /AD /N *.$$$' .
>>
>> * Save current default folder, and change to one with R/W permission *.
>> PRESERVE.
>> CD fpFolder_With_ReadWriteRights .
>>
>> * Datasets to matrix *.
>> DATASET DECLARE results1 .
>> DATASET DECLARE results2 .
>> DATASET DECLARE results3 .
>> DATASET DECLARE results4 .
>>
>> DATASET ACTIVATE abc.
>> DATASET DISPLAY.
>>
>> MATRIX.
>>   GET abc / FILE * .
>>   COMPUTE Tabc = T(abc).
>>   SAVE Tabc / OUTFILE results1 .
>>   SAVE Tabc / OUTFILE results2 .
>> END MATRIX.
>>
>> DATASET DISPLAY.
>>
>> * Restore the initial default folder: probably the SPSS installation
>> folder *.
>> RESTORE.
>>
>> * Matrix blues *.
>> MATRIX.
>>   GET abc / FILE * .
>>   COMPUTE Tabc = T(abc).
>>   SAVE Tabc / OUTFILE results3 .
>>   SAVE Tabc / OUTFILE results4 .
>> END MATRIX.
>>
>> DATASET DISPLAY.
>>
>>  * DATASET CLOSE results1 .
>>  * DATASET CLOSE results2 .
>>  * DATASET CLOSE results3 .
>>  * DATASET CLOSE results4 .
>> *-----------------------------------------------------*.
>
>
>
>
>
> -----
> --
> Bruce Weaver

> bweaver@

>
http://sites.google.com/a/lakeheadu.ca/bweaver/
>
> "When all else fails, RTFM."
>
> NOTE: My Hotmail account is not monitored regularly.
> To send me an e-mail, please use the address shown above.
>
> --
> View this message in context:
>
http://spssx-discussion.1045642.n5.nabble.com/SERIOUS-problem-with-Multiple-datasets-in-MATRIX-procedure-tp5722765p5725745.html
>
> Sent from the SPSSX Discussion mailing list archive at Nabble.com.
>
> =====================
> To manage your subscription to SPSSX-L, send a message to

> LISTSERV@.UGA

>  (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





-----
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
--
View this message in context:
http://spssx-discussion.1045642.n5.nabble.com/SERIOUS-problem-with-Multiple-datasets-in-MATRIX-procedure-tp5722765p5725747.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (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