bugPSPP - Bugs: bug #19819, regression crashes if dependent...

 
 

bug #19819: regression crashes if dependent variable is the same as analytic variable

Submitter:  John Darrington <jmd>
Submitted:  Tue 08 May 2007 05:51:26 AM UTC
   
 
Category:  Numerical Errors Severity:  5 - Average
Status:  None Assigned to:  jstover
Open/Closed:  Closed Release:  Before 0.6.0
Effort:  0.00
* Mandatory Fields

Add a New Comment Rich Markup
   

Jump to the original submission

Thu 16 Aug 2007 03:18:28 AM UTC, comment #6: 

Except  for a typo in the word "squares", I think this is fine.

John Darrington <jmd>
Group administrator
Wed 15 Aug 2007 07:25:53 PM UTC, comment #5: 

I just checked in a fix. Let me know if it works for you. I wasn't sure how to word the warning/error pspp reports, so let me know if should change it.

Jason H Stover <jstover>
Group Member
Thu 02 Aug 2007 03:10:13 AM UTC, comment #4: 

Just to keep the discussion up-to-date:

>As an analyst using PSPP, would you find it useful to see the
>report of y=x as the least-squares line? If so, then I'd do that.


I spoke to Jason off-line and he said that this would be his preferred behavior.

Ben Pfaff <blp>
Group administrator
Mon 30 Jul 2007 03:57:21 AM UTC, comment #3: 


>The PSPP REGRESSION procedure either could fail, or it could
>report the line y=x as the least squares line and perhaps also
>print a warning. Which way should PSPP behave?


As an analyst using PSPP, would you find it useful to see the report of y=x as the least-squares line?  If so, then I'd do that.  Otherwise, I'd go with the failure option.

My suspicion is that I'd find the option of failing more useful as a user: it will take me less time to figure out what I did wrong.  But it's up to you.

Ben Pfaff <blp>
Group administrator
Tue 15 May 2007 08:56:57 PM UTC, comment #2: 

The REGRESSION procedure in SPSS will fail under the syntax filed with this bug. It prints an error which states only one variable was specified in the VARIABLE subcommand, then stops.

So we have a choice: The PSPP REGRESSION procedure either could fail, or it could report the line y=x as the least squares line and perhaps also print a warning. Which way should PSPP behave?

Jason H Stover <jstover>
Group Member
Tue 08 May 2007 03:00:05 PM UTC, comment #1: 

I just checked in a fix that keeps pspp from crashing. It's only a partial fix, though. If the explanatory and dependent variables are the same, then the least-squares line should be the line y=x, but that isn't what the coefficients table says, so I will have to fix that. But the ANOVA table looks correct now, and R-squared is 1.

Jason H Stover <jstover>
Group Member
Tue 08 May 2007 05:51:26 AM UTC, original submission:  

This code causes a segfault:

DATA LIST LIST /x y .
BEGIN DATA.
70  .92 
50 .85
80 .84
20 .84
END DATA

REGRESSION /var = x
        /dependent x.

John Darrington <jmd>
Group administrator

 

(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 (Updated the item)
  • -email is unavailable- added by jstover (Posted a comment)
  • -email is unavailable- added by jmd (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
    2007-08-16 jstover Open/ClosedOpen Closed
    2007-08-16 jstover Assigned tojmd jstover
    2007-07-25 blp ReleaseNone Before 0.6.0
    2007-05-15 jstover Assigned tojstover jmd

    Back to the top

    Powered by Savane 3.13-02a9.
    Corresponding source code