Datasets or files?

[Robert L]

In my own work, I try to avoid saving data files until when it is necessary. Instead the use of datasets, copying, activating and such, has been useful and convenient. However, it has not been that simple to convince others that they should try the same approach. One reason might be that most of them are stuck with menus and not syntax, and since they all are accustomed to dealing with files in most other systems, it seems to follow that the most natural approach is to use files in SPSS too.

But I think there might be more conceptual issues as well. Are there any analogies or other ways which could make it easier for SPSS users to come to grips with datasets and files, and possibly also to make them consider using datasets rather than files? Suggestions are most welcome.

Robert

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

[Jon Peck]

Datasets are actually temporary files, but they disappear automatically when closed or have their names removed and become inactive.  So they reduce clutter, and minimize the confusion that might arise when a bunch of intermediate files are left lying around.  Since a dataset does not have a visible file location, they can also reduce the location dependencies in a batch of syntax.  Datasets were designed to facilitate working with multiple sets of data where they all need to remain open even though only one is actually active at a time.  Simply having two Data Editor windows visible at the same time can be a great convenience (although sometimes also confusing).

There may also be some performance benefits to dataset, because the internal compression process is a little different, but these are small.

On Tue, Mar 22, 2016 at 1:56 AM, Robert Lundqvist <[hidden email]> wrote:

In my own work, I try to avoid saving data files until when it is necessary. Instead the use of datasets, copying, activating and such, has been useful and convenient. However, it has not been that simple to convince others that they should try the same approach. One reason might be that most of them are stuck with menus and not syntax, and since they all are accustomed to dealing with files in most other systems, it seems to follow that the most natural approach is to use files in SPSS too.

But I think there might be more conceptual issues as well. Are there any analogies or other ways which could make it easier for SPSS users to come to grips with datasets and files, and possibly also to make them consider using datasets rather than files? Suggestions are most welcome.

Robert

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

[Turner, John E. (VADOC)]

I use a temporary file to save files between major operations.  I perform a lot of data exploratory work and it helps to be able to go back to a step in the process rather than start over.  I believe when you are performing a lot of Sorts & Merges of large data sets, it helps to have a convenient place to go back to.  Otherwise, you have to take a coffee break while your entire syntax file runs again.

Just a thought from the dark side of data analysis.

JT

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Robert Lundqvist
Sent: Tuesday, March 22, 2016 3:56 AM
To: [hidden email]
Subject: Datasets or files?

In my own work, I try to avoid saving data files until when it is necessary. Instead the use of datasets, copying, activating and such, has been useful and convenient. However, it has not been that simple to convince others that they should try the same approach. One reason might be that most of them are stuck with menus and not syntax, and since they all are accustomed to dealing with files in most other systems, it seems to follow that the most natural approach is to use files in SPSS too.

But I think there might be more conceptual issues as well. Are there any analogies or other ways which could make it easier for SPSS users to come to grips with datasets and files, and possibly also to make them consider using datasets rather than files? Suggestions are most welcome.

Robert

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