Prev: Non-linear curve fit
Next: How to combine cellfun and strfind?
From: Marc on 2 Jun 2010 10:08
Hello,

is there a was to use regstats with lscov? Apparently the x2fx function may enable me to do so... anyone has a clue?`

Thanks in advance

Marc
From: Peter Perkins on 2 Jun 2010 12:52
On 6/2/2010 10:08 AM, Marc wrote:
> Hello,
>
> is there a was to use regstats with lscov? Apparently the x2fx function
> may enable me to do so... anyone has a clue?`

Marc, they both do linear regression, and REGSTATS computes everything
that LSCOV does (and lots more), so perhaps you are asking how to get
the outputs of REGSTATS with a weighted regression, or with general
linear model (i.e., error covariance matrix known up to a scale factor)?
From: Marc on 9 Jun 2010 03:01
Thank you Peter for your answer.

I was going through a script that uses the LSCOV function to compute an exponentially weighted regression and wanted to take advantage of the built in summary stats output of regstats without having to code them myself...

I am back on this topic today, will update the thread if I find anything.

Best Regards

Marc

Peter Perkins <Peter.Perkins(a)MathRemoveThisWorks.com> wrote in message <hu6299$97g$1(a)fred.mathworks.com>...
> On 6/2/2010 10:08 AM, Marc wrote:
> > Hello,
> >
> > is there a was to use regstats with lscov? Apparently the x2fx function
> > may enable me to do so... anyone has a clue?`
>
> Marc, they both do linear regression, and REGSTATS computes everything
> that LSCOV does (and lots more), so perhaps you are asking how to get
> the outputs of REGSTATS with a weighted regression, or with general
> linear model (i.e., error covariance matrix known up to a scale factor)?
 | 
Pages: 1
Prev: Non-linear curve fit
Next: How to combine cellfun and strfind?