bugPSPP - Bugs: bug #46727, REGRESSION subcommand SAVE does...

 
 

bug #46727: REGRESSION subcommand SAVE does not honor filter variables

Submitter:  elias tsolis <estatistics>
Submitted:  Sun 20 Dec 2015 04:21:13 PM UTC
   
 
Category:  Syntax Parser Severity:  5 - Average
Status:  Confirmed Assigned to:  None
Open/Closed:  Open Release:  None
Effort:  0.00
* Mandatory Fields

Add a New Comment Rich Markup
   

Jump to the original submission

Fri 05 Feb 2016 03:08:41 PM UTC, comment #6: 

Downgrading, since Ben pushed a partial fix.

John Darrington <jmd>
Group administrator
Sun 03 Jan 2016 06:28:49 AM UTC, comment #5: 

I added the warning.

Ben Pfaff <blp>
Group administrator
Sun 03 Jan 2016 06:23:29 AM UTC, comment #4: 

Elias, I can confirm that filters are ignored.  Before we can change that, we need to define the desired behavior for the filtered cases.  Should the new variables be system-missing for the filtered cases?  Or should the new variables be calculated as they are for other cases, even though their values did not participate in calculating the regression parameters?

John, I don't think this is related to bug #44877 (which I know I said I'd fix; I still plan to).

For the moment, I'll aim to add a warning if FILTER is present when REGRESSION with a SAVE subcommand is processed.

Ben Pfaff <blp>
Group administrator
Mon 21 Dec 2015 10:22:20 AM UTC, comment #3: 

It seems you are right.  The /SAVE command behaves badly.  This
is probably related to bug #44877 which is still not fixed.

John Darrington <jmd>
Group administrator
Sun 20 Dec 2015 05:11:34 PM UTC, comment #2: 

 After a lot of experimentation,
I found out that Regression command IGNORES ANY FILTER variable when "/SAVE= RESID." subcommand is used for unknown reason!!!

Can anyone repeat this bug?

I use the following syntax:


REGRESSION /VARIABLES= var1 var2
   /DEPENDENT= var0
   /STATISTICS=COEFF R ANOVA
   /SAVE= RESID.

elias tsolis <estatistics>
Sun 20 Dec 2015 04:46:40 PM UTC, comment #1: 

The only way to make it work is the following one:

Example:
Computing a filter variable based on a cut-off score of a variable (line 1), and then, you apply the filter (line 2), and finally, you compute a new variable based on the old one (the one that is predicted by vars - line three).

A Work around to this bug!

compute filter2=RES2<60
       FILTER BY filter2.
      compute numall2=numall

elias tsolis <estatistics>
Sun 20 Dec 2015 04:21:13 PM UTC, original submission:  

Filter variable is not working for regression!!!

I have a filter variable 0/1 and when i produce regression statistics together with plots from examine command, i get TWO very different results based on sampel size!

Any adjustment of Filter variable has no any impact on sample size used by regression command!!!

Check it out!


elias tsolis <estatistics>

 

(Note: upload size limit is set to 16384 kB, after insertion of the required escape characters.)

Attach Files:
   
   
Comment:
   

No files currently attached

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -email is unavailable- added by blp (Posted a comment)
  • -email is unavailable- added by jmd (Posted a comment)
  • -email is unavailable- added by estatistics (Submitted the item)
  •  

    There are 0 votes so far. Votes easily highlight which items people would like to see resolved in priority, independently of the priority of the item set by tracker managers.

    Only logged-in users can vote.

     

    Follow 4 latest changes.

    Date Changed by Updated Field Previous Value => Replaced by
    2016-02-05 jmd Severity7 - Major 5 - Average
    2016-01-03 blp SummaryFilter variable is not workign for regression!!! REGRESSION subcommand SAVE does not honor filter variables
    2015-12-21 jmd Severity5 - Average 7 - Major
        StatusNone Confirmed

    Back to the top

    Powered by Savane 3.13-f8d8.
    Corresponding source code