f



SPSS 11.5.2.1 Hangs with "xsave", Warning #552, Unrecoverable error at 0x64f0075f, writing to 0x00000004 :-(

I use a procedure to create a similarity matrix from raw data, in order 
to use that proximity matrix in a MDS and in a cluster analysis (in 
order to do a "concept mapping").

The procedure works well on SPSS 10 and SPSS 12. However, SPSS 11.5.2.1 
(my version of SPSS) hangs when doing the first xsave (8th line) [see 
title of the message for exact error).

vector it = pomme to cyan.
vector hit (12, F4).
loop #j = 1 to 12.
+ compute item = #j.
+ loop #k = 1 to 12.
+ compute hit(#k) = (it(#j) = it(#k)).
+ end loop.
+ xsave outfile = hits.sav
    /keep = id item hit1 to hit12 .
end loop.
execute.

get file = hits.sav .

aggregate outfile = *
	/break = item
	/pomme citron cannelle sel poivre jambon saucisso salami bleu azur 
magenta cyan = sum(hit1 to hit12).

string ROWTYPE_ (A8) .
COMPUTE ROWTYPE_ = 'PROX' .
save outfile = qsmat.sav /keep = ROWTYPE_ pomme to cyan .

FLIP
  VARIABLES= pomme to cyan .
rename variables (case_lbl = VARNAME_) .
match files /file = * /file = qsmat.sav / drop = var001 to var012 .

execute.

save outfile = qsmat_1.sav 	
    /keep = ROWTYPE_ VARNAME_ pomme to cyan .

get file = qsmat_1.sav .


SPSS support was of no help :

"Although we offer operational (how to) support for previous releases of 
SPSS. We only offer full support for SPSS 13. This issue you have 
brought does not occur in SPSS 12 or 13, we cannot pursue 11.5 issues 
further.

However I will suggest that you could free up as much space on the 
clients hard drive and SPSS Temp directory. The 0xf... error is an 
indicator of low system resources."

....I have 14gb left on my HD... ?!

This issue has been described elsewhere:
http://www.listserv.uga.edu/cgi-bin/wa?A2=ind0306&L=spssx-l&F=&S=&P=35481

Is there any "PROGRAMMATION" solution (using another command than xsave) 
? i.e. without reinstalling SPSS ?

Many thanks in advance for any suggestion,

michael
0
Michael
10/29/2004 9:58:07 PM
comp.soft-sys.stat.spss 5679 articles. 0 followers. Post Follow

4 Replies
4652 Views

Similar Articles

[PageSpeed] 8

Hi Michael,

I had a similar problem using XSAVE in 11.5.x a while ago. It was 
suggested that I put a CACHE command in the syntax just before the XSAVE 
loop. That fixed the problem for me, so perhaps it will help here too.

Cheers,
Kylie.


Michael Cantinotti wrote:
> I use a procedure to create a similarity matrix from raw data, in order 
> to use that proximity matrix in a MDS and in a cluster analysis (in 
> order to do a "concept mapping").
> 
> The procedure works well on SPSS 10 and SPSS 12. However, SPSS 11.5.2.1 
> (my version of SPSS) hangs when doing the first xsave (8th line) [see 
> title of the message for exact error).
> 
> vector it = pomme to cyan.
> vector hit (12, F4).
> loop #j = 1 to 12.
> + compute item = #j.
> + loop #k = 1 to 12.
> + compute hit(#k) = (it(#j) = it(#k)).
> + end loop.
> + xsave outfile = hits.sav
>    /keep = id item hit1 to hit12 .
> end loop.
> execute.
> 
> get file = hits.sav .
> 
> aggregate outfile = *
>     /break = item
>     /pomme citron cannelle sel poivre jambon saucisso salami bleu azur 
> magenta cyan = sum(hit1 to hit12).
> 
> string ROWTYPE_ (A8) .
> COMPUTE ROWTYPE_ = 'PROX' .
> save outfile = qsmat.sav /keep = ROWTYPE_ pomme to cyan .
> 
> FLIP
>  VARIABLES= pomme to cyan .
> rename variables (case_lbl = VARNAME_) .
> match files /file = * /file = qsmat.sav / drop = var001 to var012 .
> 
> execute.
> 
> save outfile = qsmat_1.sav    
>    /keep = ROWTYPE_ VARNAME_ pomme to cyan .
> 
> get file = qsmat_1.sav .
> 
> 
> SPSS support was of no help :
> 
> "Although we offer operational (how to) support for previous releases of 
> SPSS. We only offer full support for SPSS 13. This issue you have 
> brought does not occur in SPSS 12 or 13, we cannot pursue 11.5 issues 
> further.
> 
> However I will suggest that you could free up as much space on the 
> clients hard drive and SPSS Temp directory. The 0xf... error is an 
> indicator of low system resources."
> 
> ...I have 14gb left on my HD... ?!
> 
> This issue has been described elsewhere:
> http://www.listserv.uga.edu/cgi-bin/wa?A2=ind0306&L=spssx-l&F=&S=&P=35481
> 
> Is there any "PROGRAMMATION" solution (using another command than xsave) 
> ? i.e. without reinstalling SPSS ?
> 
> Many thanks in advance for any suggestion,
> 
> michael
0
Kylie
10/30/2004 3:57:59 AM
Hi Kylie,

thanks for your reply. I tried to insert the command "CACHE." after the 
third line (1) (see below) or after the seventh line (2) (see below).

However, I keep getting the same bloody errror.

Do you think I tried the correct way ? i.e. is it like you intended ?

Regards,

Michael

Kylie Lange wrote:
> Hi Michael,
> 
> I had a similar problem using XSAVE in 11.5.x a while ago. It was 
> suggested that I put a CACHE command in the syntax just before the XSAVE 
> loop. That fixed the problem for me, so perhaps it will help here too.
> 
> Cheers,
> Kylie.
> 
> 
> Michael Cantinotti wrote:
> 
>> I use a procedure to create a similarity matrix from raw data, in 
>> order to use that proximity matrix in a MDS and in a cluster analysis 
>> (in order to do a "concept mapping").
>>
>> The procedure works well on SPSS 10 and SPSS 12. However, SPSS 
>> 11.5.2.1 (my version of SPSS) hangs when doing the first xsave (8th 
>> line) [see title of the message for exact error).
>>
>> vector it = pomme to cyan.
>> vector hit (12, F4).
>> loop #j = 1 to 12.
HERE (1)
>> + compute item = #j.
>> + loop #k = 1 to 12.
>> + compute hit(#k) = (it(#j) = it(#k)).
>> + end loop.
HERE (2)
>> + xsave outfile = hits.sav
>>    /keep = id item hit1 to hit12 .
>> end loop.
>> execute.
>>
>> get file = hits.sav .
>>
>> aggregate outfile = *
>>     /break = item
>>     /pomme citron cannelle sel poivre jambon saucisso salami bleu azur 
>> magenta cyan = sum(hit1 to hit12).
>>
>> string ROWTYPE_ (A8) .
>> COMPUTE ROWTYPE_ = 'PROX' .
>> save outfile = qsmat.sav /keep = ROWTYPE_ pomme to cyan .
>>
>> FLIP
>>  VARIABLES= pomme to cyan .
>> rename variables (case_lbl = VARNAME_) .
>> match files /file = * /file = qsmat.sav / drop = var001 to var012 .
>>
>> execute.
>>
>> save outfile = qsmat_1.sav       /keep = ROWTYPE_ VARNAME_ pomme to 
>> cyan .
>>
>> get file = qsmat_1.sav .
>>
>>
>> SPSS support was of no help :
>>
>> "Although we offer operational (how to) support for previous releases 
>> of SPSS. We only offer full support for SPSS 13. This issue you have 
>> brought does not occur in SPSS 12 or 13, we cannot pursue 11.5 issues 
>> further.
>>
>> However I will suggest that you could free up as much space on the 
>> clients hard drive and SPSS Temp directory. The 0xf... error is an 
>> indicator of low system resources."
>>
>> ...I have 14gb left on my HD... ?!
>>
>> This issue has been described elsewhere:
>> http://www.listserv.uga.edu/cgi-bin/wa?A2=ind0306&L=spssx-l&F=&S=&P=35481
>>
>> Is there any "PROGRAMMATION" solution (using another command than 
>> xsave) ? i.e. without reinstalling SPSS ?
>>
>> Many thanks in advance for any suggestion,
>>
>> michael
0
Michael
11/1/2004 4:29:24 PM
Hi Michael,

I was going on memory, so I've dug out the code that was I using 
(below). Without the SAVE OUTFILE and CAHCE commands the XSAVE LOOP hung 
indefinitely and SPSS needed to be killed through the task manager (this 
was using SPSS 11.5.2.1 on Windows NT).

Another thing you could try - on AnswerNet the following fix was posted 
(again, referring to SPSS hanging rather than raising the error message 
you're getting but it may help).

=====================================

AnswerNet No. 32938:
I have updated to the SPSS 11.5.2.x for Windows via the patch on the 
SPSS web site. When I run a command syntax job that contains an XSAVE 
command in it, SPSS seems to hang on processing that command and the job 
never completes. I was able to run this job successfully in previous 
versions of SPSS for Windows. Is there a fix for this problem?

Resolution Description:
A fix is available on the SPSS FTP site for download:
ftp://ftp.spss.com/pub/spss/windows
If prompted for a user name and password, use anonymous for the user 
name and a valid e-mail address for the password. Please download 
spsswfil.dll into your SPSS directory and overwrite the existing file.

=====================================

DATA LIST LIST /sla(F8.2) p1(F8.1) p2(F8.1) p3(F8.1) p4(F8.1)
nb(F8.0).
BEGIN DATA
10012 0 1 0 0 1
10025 .5 0 .5 0 3
10045 .4 .3 0 .3 4
10055 0 0 0 1 1
END DATA.

SAVE OUTFILE='c:\temp\temp1.sav'.
CACHE.

LOOP cnt=1 TO nb.
XSAVE OUTFILE='c:\temp\temp2.sav' /KEEP sla TO p4.
END LOOP.
EXE.
GET FILE='c:\temp\temp2.sav'.

=====================================

Hope this helps.

Cheers,
Kylie.


Michael Cantinotti wrote:

> Hi Kylie,
> 
> thanks for your reply. I tried to insert the command "CACHE." after the 
> third line (1) (see below) or after the seventh line (2) (see below).
> 
> However, I keep getting the same bloody errror.
> 
> Do you think I tried the correct way ? i.e. is it like you intended ?
> 
> Regards,
> 
> Michael
> 
> Kylie Lange wrote:
> 
>> Hi Michael,
>>
>> I had a similar problem using XSAVE in 11.5.x a while ago. It was 
>> suggested that I put a CACHE command in the syntax just before the 
>> XSAVE loop. That fixed the problem for me, so perhaps it will help 
>> here too.
>>
>> Cheers,
>> Kylie.
>>
>>
>> Michael Cantinotti wrote:
>>
>>> I use a procedure to create a similarity matrix from raw data, in 
>>> order to use that proximity matrix in a MDS and in a cluster analysis 
>>> (in order to do a "concept mapping").
>>>
>>> The procedure works well on SPSS 10 and SPSS 12. However, SPSS 
>>> 11.5.2.1 (my version of SPSS) hangs when doing the first xsave (8th 
>>> line) [see title of the message for exact error).
>>>
>>> vector it = pomme to cyan.
>>> vector hit (12, F4).
>>> loop #j = 1 to 12.
> 
> HERE (1)
> 
>>> + compute item = #j.
>>> + loop #k = 1 to 12.
>>> + compute hit(#k) = (it(#j) = it(#k)).
>>> + end loop.
> 
> HERE (2)
> 
>>> + xsave outfile = hits.sav
>>>    /keep = id item hit1 to hit12 .
>>> end loop.
>>> execute.
>>>
>>> get file = hits.sav .
>>>
>>> aggregate outfile = *
>>>     /break = item
>>>     /pomme citron cannelle sel poivre jambon saucisso salami bleu 
>>> azur magenta cyan = sum(hit1 to hit12).
>>>
>>> string ROWTYPE_ (A8) .
>>> COMPUTE ROWTYPE_ = 'PROX' .
>>> save outfile = qsmat.sav /keep = ROWTYPE_ pomme to cyan .
>>>
>>> FLIP
>>>  VARIABLES= pomme to cyan .
>>> rename variables (case_lbl = VARNAME_) .
>>> match files /file = * /file = qsmat.sav / drop = var001 to var012 .
>>>
>>> execute.
>>>
>>> save outfile = qsmat_1.sav       /keep = ROWTYPE_ VARNAME_ pomme to 
>>> cyan .
>>>
>>> get file = qsmat_1.sav .
>>>
>>>
>>> SPSS support was of no help :
>>>
>>> "Although we offer operational (how to) support for previous releases 
>>> of SPSS. We only offer full support for SPSS 13. This issue you have 
>>> brought does not occur in SPSS 12 or 13, we cannot pursue 11.5 issues 
>>> further.
>>>
>>> However I will suggest that you could free up as much space on the 
>>> clients hard drive and SPSS Temp directory. The 0xf... error is an 
>>> indicator of low system resources."
>>>
>>> ...I have 14gb left on my HD... ?!
>>>
>>> This issue has been described elsewhere:
>>> http://www.listserv.uga.edu/cgi-bin/wa?A2=ind0306&L=spssx-l&F=&S=&P=35481 
>>>
>>>
>>> Is there any "PROGRAMMATION" solution (using another command than 
>>> xsave) ? i.e. without reinstalling SPSS ?
>>>
>>> Many thanks in advance for any suggestion,
>>>
>>> michael
0
Kylie
11/2/2004 12:01:37 AM
Dear Kylie,

I really appreciate your help. After I copied "spsswfil.dll" in my SPSS 
11.5.2.1 directory, XSAVE works now !

I feel a little bit disappointed with SPSS support. They have the 
solution for the problem and don't give it to their customers. What kind 
of customer support is this ??? SAS technicians are more serious...

Fortunately, there is this forum and people, like you, who are sharing 
their knowledge.

Thanks again !

Regards,

Michael

Kylie Lange wrote:
> 
> Hi Michael,
> 
> I was going on memory, so I've dug out the code that was I using 
> (below). Without the SAVE OUTFILE and CAHCE commands the XSAVE LOOP hung 
> indefinitely and SPSS needed to be killed through the task manager (this 
> was using SPSS 11.5.2.1 on Windows NT).
> 
> Another thing you could try - on AnswerNet the following fix was posted 
> (again, referring to SPSS hanging rather than raising the error message 
> you're getting but it may help).
> 
> =====================================
> 
> AnswerNet No. 32938:
> I have updated to the SPSS 11.5.2.x for Windows via the patch on the 
> SPSS web site. When I run a command syntax job that contains an XSAVE 
> command in it, SPSS seems to hang on processing that command and the job 
> never completes. I was able to run this job successfully in previous 
> versions of SPSS for Windows. Is there a fix for this problem?
> 
> Resolution Description:
> A fix is available on the SPSS FTP site for download:
> ftp://ftp.spss.com/pub/spss/windows
> If prompted for a user name and password, use anonymous for the user 
> name and a valid e-mail address for the password. Please download 
> spsswfil.dll into your SPSS directory and overwrite the existing file.
> 
> =====================================
> 
> DATA LIST LIST /sla(F8.2) p1(F8.1) p2(F8.1) p3(F8.1) p4(F8.1)
> nb(F8.0).
> BEGIN DATA
> 10012 0 1 0 0 1
> 10025 .5 0 .5 0 3
> 10045 .4 .3 0 .3 4
> 10055 0 0 0 1 1
> END DATA.
> 
> SAVE OUTFILE='c:\temp\temp1.sav'.
> CACHE.
> 
> LOOP cnt=1 TO nb.
> XSAVE OUTFILE='c:\temp\temp2.sav' /KEEP sla TO p4.
> END LOOP.
> EXE.
> GET FILE='c:\temp\temp2.sav'.
> 
> =====================================
> 
> Hope this helps.
> 
> Cheers,
> Kylie.
> 
> 
> Michael Cantinotti wrote:
> 
>> Hi Kylie,
>>
>> thanks for your reply. I tried to insert the command "CACHE." after 
>> the third line (1) (see below) or after the seventh line (2) (see below).
>>
>> However, I keep getting the same bloody errror.
>>
>> Do you think I tried the correct way ? i.e. is it like you intended ?
>>
>> Regards,
>>
>> Michael
>>
>> Kylie Lange wrote:
>>
>>> Hi Michael,
>>>
>>> I had a similar problem using XSAVE in 11.5.x a while ago. It was 
>>> suggested that I put a CACHE command in the syntax just before the 
>>> XSAVE loop. That fixed the problem for me, so perhaps it will help 
>>> here too.
>>>
>>> Cheers,
>>> Kylie.
>>>
>>>
>>> Michael Cantinotti wrote:
>>>
>>>> I use a procedure to create a similarity matrix from raw data, in 
>>>> order to use that proximity matrix in a MDS and in a cluster 
>>>> analysis (in order to do a "concept mapping").
>>>>
>>>> The procedure works well on SPSS 10 and SPSS 12. However, SPSS 
>>>> 11.5.2.1 (my version of SPSS) hangs when doing the first xsave (8th 
>>>> line) [see title of the message for exact error).
>>>>
>>>> vector it = pomme to cyan.
>>>> vector hit (12, F4).
>>>> loop #j = 1 to 12.
>>
>>
>> HERE (1)
>>
>>>> + compute item = #j.
>>>> + loop #k = 1 to 12.
>>>> + compute hit(#k) = (it(#j) = it(#k)).
>>>> + end loop.
>>
>>
>> HERE (2)
>>
>>>> + xsave outfile = hits.sav
>>>>    /keep = id item hit1 to hit12 .
>>>> end loop.
>>>> execute.
>>>>
>>>> get file = hits.sav .
>>>>
>>>> aggregate outfile = *
>>>>     /break = item
>>>>     /pomme citron cannelle sel poivre jambon saucisso salami bleu 
>>>> azur magenta cyan = sum(hit1 to hit12).
>>>>
>>>> string ROWTYPE_ (A8) .
>>>> COMPUTE ROWTYPE_ = 'PROX' .
>>>> save outfile = qsmat.sav /keep = ROWTYPE_ pomme to cyan .
>>>>
>>>> FLIP
>>>>  VARIABLES= pomme to cyan .
>>>> rename variables (case_lbl = VARNAME_) .
>>>> match files /file = * /file = qsmat.sav / drop = var001 to var012 .
>>>>
>>>> execute.
>>>>
>>>> save outfile = qsmat_1.sav       /keep = ROWTYPE_ VARNAME_ pomme to 
>>>> cyan .
>>>>
>>>> get file = qsmat_1.sav .
>>>>
>>>>
>>>> SPSS support was of no help :
>>>>
>>>> "Although we offer operational (how to) support for previous 
>>>> releases of SPSS. We only offer full support for SPSS 13. This issue 
>>>> you have brought does not occur in SPSS 12 or 13, we cannot pursue 
>>>> 11.5 issues further.
>>>>
>>>> However I will suggest that you could free up as much space on the 
>>>> clients hard drive and SPSS Temp directory. The 0xf... error is an 
>>>> indicator of low system resources."
>>>>
>>>> ...I have 14gb left on my HD... ?!
>>>>
>>>> This issue has been described elsewhere:
>>>> http://www.listserv.uga.edu/cgi-bin/wa?A2=ind0306&L=spssx-l&F=&S=&P=35481 
>>>>
>>>>
>>>> Is there any "PROGRAMMATION" solution (using another command than 
>>>> xsave) ? i.e. without reinstalling SPSS ?
>>>>
>>>> Many thanks in advance for any suggestion,
>>>>
>>>> michael
0
Michael
11/2/2004 3:36:45 PM
Reply: