SPSS processor terminated due to unrecoverable error at . . .

[Chris Poole-2]

Dear List,

I am running SPSS v15.0.1 on Windows XP (SP3) performing some time-dependent Cox survival analysis on a modest dataset (n=30,000 cases).

I have performed the same analysis many times without problem yet today I am getting the message listed in the title - "SPSS processor terminated due to unrecoverable error at . . . - fetching . . . "

I've tried increasing the page file size for memory allocation and set spsswin.exe to use only CPU1 as per instructions from users with a similar problem to no avail.

Any ideas?

Kind regards,

Chris.

--
Dr Chris Poole
Pharmatelligence LLP
Pharma Research Centre | MediCentre | Cardiff | CF14 4UJ | UK
[hidden email]
P. [phone] | F. [phone] | M. [phone]

[Mark A Davenport MADAVENP]

I have a dataset with multiple rows per ID.  There are five columns: 1 for the ID and four string vars (two have complete data, two have some missing data).

ID                  STR1      STR2     STR3     STR4

107                                                   zzz            aaa
107                                                   zzz            aaa
345                xxx            yyy          zzz            aaa
345                xxx            yyy          zzz            aaa                
345                xxx            yyy          zzz            aaa
546                xxx            yyy          zzz            aaa
etc.

 I want to reduce this dataset to a set with a single row per ID and the four string vars attached at the end.

Using:

SORT CASES BY ID .
CASESTOVARS
  /ID=ID
  /GROUPBY=INDEX
  /COUNT=NOEUS "".

This is a very simple and straight forward issue of cases to vars.  I have done this a thousand times.  however, I am unable to get STR1 and STR2
to break out.  What happens is I get a new single row (as expected) wherein the original first row's STR1 and STR2 values appear in the new dataset in the SRT1.1 and STR1.2 place and SRT1.3 and STR1.4 break out as expected.    Does CASESTOVARS require complete data?  I don't recall reading anything about such a requirement.

mark


***************************************************************************************************************************************************************
Mark A. Davenport Ph.D.
Senior Research Analyst
Office of Institutional Research
The University of North Carolina at Greensboro
336.256.0395
[hidden email]

'An approximate answer to the right question is worth a good deal more than an exact answer to an approximate question.' --a paraphrase of J. W. Tukey (1962)

[Maguin, Eugene]

Mark,

I'm confused

Wouldn't you expect to see

ID   STR1.1  STR1.2  STR1.3  STR2.1  STR2.2  STR2.3  STR3  STR4
107                                                   zzz   aaa
345     xxx                     yyy                   zzz   aaa
546     xxx                     yyy                   zzz   aaa


STR3 and STR4 are fixed, so no new variables for those two. Case 345 has
three records so that defines new .1, .2, and .3 variables for those
variables with not fixed values. You must have a case somewhere with four
rows to get a new .4 variable.

Gene Maguin


>>I have a dataset with multiple rows per ID.  There are five columns: 1 for
the ID and four string vars (two have complete data, two have some missing
data).

ID                  STR1      STR2     STR3     STR4
107                                                   zzz            aaa
107                                                   zzz            aaa
345                xxx            yyy          zzz            aaa
345                xxx            yyy          zzz            aaa

345                xxx            yyy          zzz            aaa
546                xxx            yyy          zzz            aaa
etc.

 I want to reduce this dataset to a set with a single row per ID and the
four string vars attached at the end.

Using:

SORT CASES BY ID .
CASESTOVARS
  /ID=ID
  /GROUPBY=INDEX
  /COUNT=NOEUS "".

This is a very simple and straight forward issue of cases to vars.  I have
done this a thousand times.  however, I am unable to get STR1 and STR2
to break out.  What happens is I get a new single row (as expected) wherein
the original first row's STR1 and STR2 values appear in the new dataset in
the SRT1.1 and STR1.2 place and SRT1.3 and STR1.4 break out as expected.
Does CASESTOVARS require complete data?  I don't recall reading anything
about such a requirement.

mark


****************************************************************************
****************************************************************************
*******
Mark A. Davenport Ph.D.
Senior Research Analyst
Office of Institutional Research
The University of North Carolina at Greensboro
336.256.0395
[hidden email]

'An approximate answer to the right question is worth a good deal more than
an exact answer to an approximate question.' --a paraphrase of J. W. Tukey
(1962)

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