New commands on the Extension Hub

[Jon Peck]

Recently I wrote about the STATS CONVERT PYTHON extension command that converts Python 2 code to Python 3 so that you can use it with Statistics 27, which does not support Python 2 out of the box.  I wrote about this on the IBM Community site here

https://community.ibm.com/community/user/datascience/viewdocument/python-2-to-python-3-conversion-for-1?CommunityKey=886b6874-0fb1-402c-8243-c70ef8179a99&tab=librarydocuments

This command is now on the Extension Hub and can be installed via the Extensions > Extension Hub menu in Statistics 27.

Commands in the title of this post is plural, so I want to report that there is another new command also on the Extension Hub: STATS RESIDUAL BOXPLOTS.  This command can be useful in checking for heteroscedasticity and functional form problems in regression.  It produces a set of linked boxplots of residuals against predicted values.  Like most extensions, it has a custom dialog box and SPSS-style syntax.  When you install it, a short article on usage is saved in the STATS_RESIDUAL_BOXPLOTS directory under the installation location.

The command requires the appropriate version of R and the R Essentials for Statistics.  While it was developed using Statistics V27, I expect that it will work with V24 and later.

--

===================== 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]

Thanks for the update, Jon.  

At the risk of setting the cat among the pigeons, if IBM had even a little
bit of interest in writing native SPSS code for the methods that currently
require R or Python extension commands, this would not be an issue.  :-|  



Jon Peck wrote

> Recently I wrote about the STATS CONVERT PYTHON extension command that
> converts Python 2 code to Python 3 so that you can use it with Statistics
> 27, which does not support Python 2 out of the box.  I wrote about this on
> the IBM Community site here
> https://community.ibm.com/community/user/datascience/viewdocument/python-2-to-python-3-conversion-for-1?CommunityKey=886b6874-0fb1-402c-8243-c70ef8179a99&tab=librarydocuments
>
> This command is now on the Extension Hub and can be installed via the
> Extensions > Extension Hub menu in Statistics 27.
>
> Commands in the title of this post is plural, so I want to report that
> there is another new command also on the Extension Hub: STATS RESIDUAL
> BOXPLOTS.  This command can be useful in checking for heteroscedasticity
> and functional form problems in regression.  It produces a set of linked
> boxplots of residuals against predicted values.  Like most extensions, it
> has a custom dialog box and SPSS-style syntax.  When you install it, a
> short article on usage is saved in the STATS_RESIDUAL_BOXPLOTS directory
> under the installation location.
>
> The command requires the appropriate version of R and the R Essentials for
> Statistics.  While it was developed using Statistics V27, I expect that it
> will work with V24 and later.
>
>
> --
> Jon K Peck

> jkpeck@

>
> =====================
> 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.

--
Sent from: http://spssx-discussion.1045642.n5.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

[Jon Peck]

While I would love to see IBM invest more heavily in statistical methods and other things for the Statistics product, it seems to me that integrating open source methods in Python and R makes sense as there is much more to do than any one company can provide - even IBM.  IBM/SPSS provides the architectural features that allow open source to be leveraged and that can't come from the open source community along with some of the statistical and graphical work that is deemed most strategic.

Note that I don't speak for IBM and don't necessarily know what is in the works.

On Tue, Mar 16, 2021 at 2:35 PM Bruce Weaver <[hidden email]> wrote:

Thanks for the update, Jon. 

At the risk of setting the cat among the pigeons, if IBM had even a little
bit of interest in writing native SPSS code for the methods that currently
require R or Python extension commands, this would not be an issue.  :-| 



Jon Peck wrote
> Recently I wrote about the STATS CONVERT PYTHON extension command that
> converts Python 2 code to Python 3 so that you can use it with Statistics
> 27, which does not support Python 2 out of the box.  I wrote about this on
> the IBM Community site here
> https://community.ibm.com/community/user/datascience/viewdocument/python-2-to-python-3-conversion-for-1?CommunityKey=886b6874-0fb1-402c-8243-c70ef8179a99&tab=librarydocuments
>
> This command is now on the Extension Hub and can be installed via the
> Extensions > Extension Hub menu in Statistics 27.
>
> Commands in the title of this post is plural, so I want to report that
> there is another new command also on the Extension Hub: STATS RESIDUAL
> BOXPLOTS.  This command can be useful in checking for heteroscedasticity
> and functional form problems in regression.  It produces a set of linked
> boxplots of residuals against predicted values.  Like most extensions, it
> has a custom dialog box and SPSS-style syntax.  When you install it, a
> short article on usage is saved in the STATS_RESIDUAL_BOXPLOTS directory
> under the installation location.
>
> The command requires the appropriate version of R and the R Essentials for
> Statistics.  While it was developed using Statistics V27, I expect that it
> will work with V24 and later.
>
>
> --
> Jon K Peck

> jkpeck@

>
> =====================
> 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.

--
Sent from: http://spssx-discussion.1045642.n5.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