From: burcu on
Seems i have another problem. Tried the code with my data which includes 5 different values on a 800 lined col vec.
my tm is a matrix of 350x350. any idea why cant i get a 5x5 matrix?

Burcu
>
> >
> > well... ACCUMARRAY typically is used (by most CSSMers) to create transition matrices... and, unless i completely misunderstood what you need...
> >
> > one of the solutions
> >
> > v=[2,3,4,3,5,2,3,5,2,3,4,2,1].'; % <- note: col vec
> > tm=accumarray([v(1:end-1),v(2:end)],1)
> > %{
> > % tm =
> > 0 0 0 0 0 % <- eg, no transition 1->
> > 1 0 3 0 0 % <- eg, transition 2->3 occurs three times...
> > 0 0 0 2 2
> > 0 1 1 0 0
> > 0 2 0 0 0
> > %}
> >
> > us
From: us on
"burcu " <burcu102(a)hotmail.com> wrote in message <hscaqp$g44$1(a)fred.mathworks.com>...
> Seems i have another problem. Tried the code with my data which includes 5 different values on a 800 lined col vec.
> my tm is a matrix of 350x350. any idea why cant i get a 5x5 matrix?
>
> Burcu
> >
> > >
> > > well... ACCUMARRAY typically is used (by most CSSMers) to create transition matrices... and, unless i completely misunderstood what you need...
> > >
> > > one of the solutions
> > >
> > > v=[2,3,4,3,5,2,3,5,2,3,4,2,1].'; % <- note: col vec
> > > tm=accumarray([v(1:end-1),v(2:end)],1)
> > > %{
> > > % tm =
> > > 0 0 0 0 0 % <- eg, no transition 1->
> > > 1 0 3 0 0 % <- eg, transition 2->3 occurs three times...
> > > 0 0 0 2 2
> > > 0 1 1 0 0
> > > 0 2 0 0 0
> > > %}
> > >
> > > us

don't top post; it's tedious...

anyhow, remember what i said...

v=[2,3,4,3,5,2,3,5,2,3,4,2,1].'; % <- note: COLumn VECtor
% so, reshape your MxN mat into a col vec...
% note: this will accumulate start-end-points as well...

us