Re: Scatterplot

[John Darrington]

Your original question was about fitting a curve, rather than
plotting one.

Plotting an arbitrary curve would require the GGRAPH command
which we don't currently implement.

J'

On Mon, Jul 20, 2009 at 05:01:26PM -0700, David Nasatir wrote:
> Thanks, Jason.
>
> Using the syntax below, I do, indeed, obtain the desired statistics and  
> the variables PRED and RESID are created.  Now, how do I get the  
> scatterplot using PRED and RES without leaving PSPP?
>
> Thanks again,
>
> David
>
>
> Jason Stover wrote:
>> With the REGRESSION procedure, the syntax would be something
>> like this:
>>
>> REGRESSION
>>      /VARIABLES= educ
>>      /DEPENDENT=      happy
>>      /STATISTICS=COEFF R ANOVA
>>      /SAVE= PRED RESID.
>>
>> You can then make scatterplots using residuals and predicted values.
>>
>> On Mon, Jul 20, 2009 at 09:42:48AM -0700, David Nasatir wrote:
>>   
>>> Hello,
>>>
>>> I'm the guy trying to fit a line to a scatterplot in PSPP. 
>>>
>>> John Darrington suggested it might be done using the REGRESSION  
>>> procedure with "...a bit more manual work".
>>>
>>> What would that work be?  What steps should I follow?
>>>
>>> I understand that the syntax in SPSS would be something like:
>>>
>>> * Curve Estimation.
>>> TSET NEWVAR=NONE .
>>> PREDICT THRU END.
>>> CURVEFIT /VARIABLES=happy  WITH educ
>>>  /CONSTANT
>>>  /MODEL=LINEAR
>>>  /PLOT FIT.
>>>
>>> but, of course, that doesn't work in the version of PSPPIRE 0.7.2 
>>> that I have been using
>>>
>>> At the moment I must move the data to Open Office Calc, do the  
>>> regression using that program and then insert a line.  I would rather 
>>> just stay in PSPPIRE 0.7.2 if possible.
>>>
>>> Thanks for your help and for all the good work you have been doing
>>>
>>> David
>>>
>>>
>>> John Darrington wrote:
>>>     
>>>> On Sun, Jul 19, 2009 at 08:37:54PM -0300, michel wrote:
>>>>         Hello,
>>>>
>>>>     
>>>>
>>>>     A Windows user is asking me if there is some way to fit a curve to a
>>>>
>>>>     scatterplot on PSPP. I don't think it is possible
>>>>
>>>>     so far, but I don't know for sure.
>>>>
>>>>     
>>>>
>>>>     Is there some way to do this?
>>>>
>>>>
>>>> Not exactly.  The procedure would be CURVEFIT, which we don't currently
>>>> implement.
>>>>
>>>> But the job could be done with the REGRESSION procedure - the user just
>>>> has to do a bit more manual work.
>>>>
>>>> J'
>>>>
>>>>   
>>>> ------------------------------------------------------------------------
>>>>
>>>> _______________________________________________
>>>> Pspp-users mailing list
>>>> address@hidden
>>>> http://lists.gnu.org/mailman/listinfo/pspp-users
>>>>        
>>>
>>> _______________________________________________
>>> Pspp-users mailing list
>>> address@hidden
>>> http://lists.gnu.org/mailman/listinfo/pspp-users
>>>     
>>
>>   
>
>
>
> _______________________________________________
> Pspp-users mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/pspp-users

-- 
PGP Public key ID: 1024D/2DE827B3 
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://pgp.mit.edu or any PGP keyserver for public key.

signature.asc
Description: Digital signature