pspp-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: PSPP multiple linear regression with Y intercept equal zero


From: John Darrington
Subject: Re: PSPP multiple linear regression with Y intercept equal zero
Date: Sat, 3 Sep 2016 22:46:55 +0200
User-agent: Mutt/1.5.23 (2014-03-12)

There are some tips on how to get missing features implemented here:
http://www.gnu.org/software/pspp/faq.html#MissingFeatures

On Sat, Sep 03, 2016 at 05:08:20PM +0000, Ether wrote:
     Thanks for replying John.
     Actually, what I want is the /ORIGIN command (force model to go through 
origin; i.e. no const term).
     I can't fake it by adding a dummy variable with all cases set to zero for 
three reasons:
     1) PSPP does not support weighted multiple linear regression, so adding a 
single dummy variable to a dataset which already contains hundreds of data 
points would not force the model's Y intercept to be zero
     2) Without weighting, it would be necessary to add hundreds (if not 
thousands) of dummy data points to force the model's Y intercept to be 
zero.???? This would be unwieldy.
     
     3) Adding hundreds (or thousands) of dummy data points would skew other 
statistical measures I am interested in.
     
     The following free statistics programs all support zero Y-intercept (no 
const term) multiple linear regression models:
     gretl
     R
     
     Octave
     
     Tanagra
     ASP 
     
     MicrOsiris 
     
     LibreOffice Calc
     
     So why don't I just use one of them instead??? Well, I have and I do.?? 
But I have several applications where I want to use PSPP; for example when 
working with students with prior exposure to SPSS.
     
     
     
           From: John Darrington <address@hidden>
      To: Ether <address@hidden> 
     Cc: "address@hidden" <address@hidden>
      Sent: Thursday, September 1, 2016 7:57 PM
      Subject: Re: PSPP multiple linear regression with Y intercept equal zero
        
     Ether,
     
     I'm not an expert on linear regression, but I think you are looking for 
the /NOORIGIN
     subcommand which unfortunately at the moment PSPP doesn't have.
     
     If this is important to you, file a "wishlist" bug and we'll ses what we 
can do.
     
     In the mean time, you can "fake it" by adding a dummy variable with all 
cases set
     to zero.
     
     J'
     
     
     On Thu, Sep 01, 2016 at 09:33:32PM +0000, Ether wrote:
     ?? ?? 
     ?? ?? Only 2 posts to the PSPP mailing list in the past 7 days.
     ?? ?? Is there a more active PSPP mailing list (or better yet, forum) 
somewhere?
     ?? ?? 
     ?? ?? 
     -- 
     Avoid eavesdropping.?? Send strong encryted email.
     PGP Public key ID: 1024D/2DE827B3 
     fingerprint = 8797 A26D 0854 2EAB 0285?? A290 8A67 719C 2DE8 27B3
     See http://sks-keyservers.net or any PGP keyserver for public key.
     
     
        

-- 
Avoid eavesdropping.  Send strong encryted email.
PGP Public key ID: 1024D/2DE827B3 
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://sks-keyservers.net or any PGP keyserver for public key.

Attachment: signature.asc
Description: Digital signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]