Re: Multiple Regression with no constant term

[ftr]

I am starting vacation tomorrow but in two weeks time I might have a job
to use with GLM. The difficulty is that I shall need to save the
residuals etc. per case.

- ftr

On 10/07/2015 20:00, John Darrington wrote:
> GLM is undocumented, because it is work in progress and not
> thoroughly tested.  
>
> However, I think the subcommands you mention are the ones that currently
> are implemented.  To be sure you'll have to look at the source.
>
>
> If you want to try out GLM and send us your feedback, it would
> be very welcome.
>
>
> J'
>
>
>
>
> On Fri, Jul 10, 2015 at 03:21:45PM +0200, ftr wrote:
>      
>      
>      On 10/07/2015 13:24, John Darrington wrote:
>      > On Fri, Jul 10, 2015 at 01:03:44PM +0200, ftr wrote:
>      >      The PSPP user guide version 0.6.2 says that the GLM command is not
>      >      implemented.
>      >      If the contrary is true all the better.
>      >      And if so, where do I find its command syntax ?
>      >      
>      > 0.6.2 is more than five years old.  If that is the version you are 
> using,
>      > I recommend that you upgrade.
>      >
>      > That said, GLM is also undocumented in 0.8.5 - it is however 
> implemented,
>      > albeit not, with all options.  It does however have the /INTERCEPT 
> subcommand.
>      > /INTERCEPT = EXCLUDE will find a model through the origin.
>      
>      I use the last version of PSPP, but the dated manual is the first that I
>      got through Google Search (from the University of Hamburg.
>      
>      If I want to use GLM which subcommands can be used ?
>      
>      I found the latest version dated 15th Feb 2015 in a news group which 
> said:
>      
>      It looks as if the sub-commands currently available are
>      
>      /MISSING={INCLUDE,EXCLUDE}
>      /INTERCEPT={INCLUDE,EXCLUDE}
>      /CRITERIA=ALPHA(n)
>      /METHOD=SSTYPE(n)
>      /DESIGN
>      /SHOWCODES
>      
>      Is this still true ?
>      
>      Is there anywhere a complete command syntax available ?
>      To use the procedure you have to look into a SPSS manual ?
>      
>      Thanks,
>      ftr
>      
>      
>      
>      _______________________________________________
>      Pspp-users mailing list
>      address@hidden
>      https://lists.gnu.org/mailman/listinfo/pspp-users
>