SPSS 15 Multiple Dataset control

[Nathaniel Lewis-20]

Hi,

Long-time user of both SPSS and the list, which is most entertaining at
times, but more importantly extremely useful in uncovering tricks to get
more out of this very powerful software.

We make the most use of it in file handling and shaping and statistical
QA of large datasets through it's automation capabilities.

I'd appreciate any advice to help me with a long time bug-bear with the
newer versions.

I can see the usefulness of having multiple data sets open at once, but
for how we use it, the file conflicts these multiple data set windows
cause when saving huge automatically built files is a major headache.

Is there any way to force SPSS (v15) to only ever use one data set
window? There may already be something there but I can't find it ;)

How would this be controlled at syntax level? I'm aware of the Dataset=
command, but this is automatically updated depending on the existing
open data sets, when you paste the syntax to a new syntax file, is there
a way of fixing this dataset number to be always dataset 1, is that one
way around this particular issue we are having?

I'm not keen on using python (but suspect that may be the only
resolution) and would like control at the syntax file level, as we
generally build our syntax files in through a java app which builds the
syntax on the fly. At that level we could control the dataset number,
but before we automate we build things by hand and then turn it into a
java supported syntax generator, and that is where my pain begins.

Thanks for your advice, and keep those tips coming!

Cheers,

Nathaniel

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

[Oliver, Richard]

At a syntax level, if you never issue a DATASET NAME or DATASET DECLARE command, then you will always be working with one datase at a time. For example:

get file 'file1.sav'.
frequencies variables=somevar.
get file 'file2.sav'.
descriptives variables=someothervar.
get data /type=xls /file='file3.xls'.
frequencies variables=all.

In the absence of any DATASET NAME commands, each unnamed dataset is automatically closed when you open the next data file.

In version 16, we added an Options setting for the GUI to provide one-file-at-a-time behavior. All this does is suppress the generation of the automatic DATASET NAME command after the file access command. In other words, if you use the GUI to open data files, the default generated syntax is something like:

get file 'filename.sav'.
dataset name dataset1 window=front.

When you select the option (in SPSS 16) to only open one file at a time, you get:

get file 'filename.sav'.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Nathaniel Lewis
Sent: Wednesday, June 18, 2008 7:28 PM
To: [hidden email]
Subject: SPSS 15 Multiple Dataset control

Hi,

Long-time user of both SPSS and the list, which is most entertaining at
times, but more importantly extremely useful in uncovering tricks to get
more out of this very powerful software.

We make the most use of it in file handling and shaping and statistical
QA of large datasets through it's automation capabilities.

I'd appreciate any advice to help me with a long time bug-bear with the
newer versions.

I can see the usefulness of having multiple data sets open at once, but
for how we use it, the file conflicts these multiple data set windows
cause when saving huge automatically built files is a major headache.

Is there any way to force SPSS (v15) to only ever use one data set
window? There may already be something there but I can't find it ;)

How would this be controlled at syntax level? I'm aware of the Dataset=
command, but this is automatically updated depending on the existing
open data sets, when you paste the syntax to a new syntax file, is there
a way of fixing this dataset number to be always dataset 1, is that one
way around this particular issue we are having?

I'm not keen on using python (but suspect that may be the only
resolution) and would like control at the syntax file level, as we
generally build our syntax files in through a java app which builds the
syntax on the fly. At that level we could control the dataset number,
but before we automate we build things by hand and then turn it into a
java supported syntax generator, and that is where my pain begins.

Thanks for your advice, and keep those tips coming!

Cheers,

Nathaniel

=======
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

[Richard Ristow]

At 08:28 PM 6/18/2008, Nathaniel Lewis wrote:

>We make the most use of SPSS in file handling, and shaping and
>statistical QA of large datasets through it's automation capabilities.
>
>I can see the usefulness of having multiple data sets open at once,
>but for how we use it, the file conflicts these multiple data set
>windows cause when saving huge automatically built files is a major headache.
>
>Is there any way to force SPSS (v15) to only ever use one data set
>window? There may already be something there but I can't find it ;)

Syntax that contains no DATASET commands should never open more than
one dataset. I believe there's no discernable difference from what
you'd see using the working file in pre-14 versions of SPSS.

The GET FILE syntax generated by the menus does include a DATASET
NAME; and a named dataset does stay open, when you load another one.
Deleting those DATASET NAME commands from your syntax should do it.

If you routinely load files using the menus, Paste the syntax instead
of running it, and delete the DATASET NAME command before running.

(An awkwardness is that, if you load from the "Recently used data"
list, there's no option to paste the syntax.)

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

[Oliver, Richard]

Minor correction:  When you check the single dataset item, you do always get a DATASET NAME command, but it always uses the same name.

-----Original Message-----
From: Oliver, Richard
Sent: Thursday, June 19, 2008 1:45 PM
To: 'Nathaniel Lewis'; [hidden email]
Subject: RE: SPSS 15 Multiple Dataset control

At a syntax level, if you never issue a DATASET NAME or DATASET DECLARE command, then you will always be working with one datase at a time. For example:

get file 'file1.sav'.
frequencies variables=somevar.
get file 'file2.sav'.
descriptives variables=someothervar.
get data /type=xls /file='file3.xls'.
frequencies variables=all.

In the absence of any DATASET NAME commands, each unnamed dataset is automatically closed when you open the next data file.

In version 16, we added an Options setting for the GUI to provide one-file-at-a-time behavior. All this does is suppress the generation of the automatic DATASET NAME command after the file access command. In other words, if you use the GUI to open data files, the default generated syntax is something like:

get file 'filename.sav'.
dataset name dataset1 window=front.

When you select the option (in SPSS 16) to only open one file at a time, you get:

get file 'filename.sav'.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Nathaniel Lewis
Sent: Wednesday, June 18, 2008 7:28 PM
To: [hidden email]
Subject: SPSS 15 Multiple Dataset control

Hi,

Long-time user of both SPSS and the list, which is most entertaining at
times, but more importantly extremely useful in uncovering tricks to get
more out of this very powerful software.

We make the most use of it in file handling and shaping and statistical
QA of large datasets through it's automation capabilities.

I'd appreciate any advice to help me with a long time bug-bear with the
newer versions.

I can see the usefulness of having multiple data sets open at once, but
for how we use it, the file conflicts these multiple data set windows
cause when saving huge automatically built files is a major headache.

Is there any way to force SPSS (v15) to only ever use one data set
window? There may already be something there but I can't find it ;)

How would this be controlled at syntax level? I'm aware of the Dataset=
command, but this is automatically updated depending on the existing
open data sets, when you paste the syntax to a new syntax file, is there
a way of fixing this dataset number to be always dataset 1, is that one
way around this particular issue we are having?

I'm not keen on using python (but suspect that may be the only
resolution) and would like control at the syntax file level, as we
generally build our syntax files in through a java app which builds the
syntax on the fly. At that level we could control the dataset number,
but before we automate we build things by hand and then turn it into a
java supported syntax generator, and that is where my pain begins.

Thanks for your advice, and keep those tips coming!

Cheers,

Nathaniel

=======
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