Prev: Delaunay/interpolation from triangular mesh
Next: Help in some matrix operations
From: James Tursa on 24 Jun 2010 16:30
And your next question, I suppose, is going to be what about an arbitrary number of 3D matrices? If that is the case, I would suggest using cell arrays and then cell2mat at the end. e.g.,

X = cell(1,1,5);
for k=1:5
X{k} = rand(10,20,k)+k;
end
C = cell2mat(X);

James Tursa
From: Kara Charaziak on 25 Jun 2010 10:45
"James Tursa" <aclassyguy_with_a_k_not_a_c(a)hotmail.com> wrote in message <i00f8s$i0$1(a)fred.mathworks.com>...
> And your next question, I suppose, is going to be what about an arbitrary number of 3D matrices? If that is the case, I would suggest using cell arrays and then cell2mat at the end. e.g.,
>
> X = cell(1,1,5);
> for k=1:5
> X{k} = rand(10,20,k)+k;
> end
> C = cell2mat(X);
>
> James Tursa

Thanks,
you are right that the way I decided to use is not very effective (especially that I load a lot of data but still should work - I need to run the whole script only one maybe two times) But I don't know how many rows the new matrix is going to have. Probably sth around 200, but it changes.

I used the cat function to merge the 3d matrices:
new_data=[];
for i=1:10
[data]=function
new_data=cat(3, new_data, data)
end
so far it seems to work (but I test my script loading only small amount of data)
the other issue I was not able to resolve was with adding descriptions to the matrix (strings)
so I want to have numbers in matrix but in each row I want to add a string indicating e.g. subject ID (need to be a string)
e.g. [20 30 1 2 3 44 5 kk-008]
are there such combined matrices? I couldn't find anything helpful so far. Maybe someone has an idea?

Thanks for all your help :)
From: us on 25 Jun 2010 11:02
"Kara Charaziak" <pink03(a)o2.pl> wrote in message <i02fe2$coa$1(a)fred.mathworks.com>...
> "James Tursa" <aclassyguy_with_a_k_not_a_c(a)hotmail.com> wrote in message <i00f8s$i0$1(a)fred.mathworks.com>...
> > And your next question, I suppose, is going to be what about an arbitrary number of 3D matrices? If that is the case, I would suggest using cell arrays and then cell2mat at the end. e.g.,
> >
> > X = cell(1,1,5);
> > for k=1:5
> > X{k} = rand(10,20,k)+k;
> > end
> > C = cell2mat(X);
> >
> > James Tursa
>
> Thanks,
> you are right that the way I decided to use is not very effective (especially that I load a lot of data but still should work - I need to run the whole script only one maybe two times) But I don't know how many rows the new matrix is going to have. Probably sth around 200, but it changes.
>
> I used the cat function to merge the 3d matrices:
> new_data=[];
> for i=1:10
> [data]=function
> new_data=cat(3, new_data, data)
> end
> so far it seems to work (but I test my script loading only small amount of data)
> the other issue I was not able to resolve was with adding descriptions to the matrix (strings)
> so I want to have numbers in matrix but in each row I want to add a string indicating e.g. subject ID (need to be a string)
> e.g. [20 30 1 2 3 44 5 kk-008]
> are there such combined matrices? I couldn't find anything helpful so far. Maybe someone has an idea?
>
> Thanks for all your help :)

again, most CSSMers would use a CELL approach (as shown above) and do the concatenation in one shot after the loop...

us
From: Kara Charaziak on 25 Jun 2010 11:33
"us " <us(a)neurol.unizh.ch> wrote in message <i02gdc$ft4$1(a)fred.mathworks.com>...
> "Kara Charaziak" <pink03(a)o2.pl> wrote in message <i02fe2$coa$1(a)fred.mathworks.com>...
> > "James Tursa" <aclassyguy_with_a_k_not_a_c(a)hotmail.com> wrote in message <i00f8s$i0$1(a)fred.mathworks.com>...
> > > And your next question, I suppose, is going to be what about an arbitrary number of 3D matrices? If that is the case, I would suggest using cell arrays and then cell2mat at the end. e.g.,
> > >
> > > X = cell(1,1,5);
> > > for k=1:5
> > > X{k} = rand(10,20,k)+k;
> > > end
> > > C = cell2mat(X);
> > >
> > > James Tursa
> >
> > Thanks,
> > you are right that the way I decided to use is not very effective (especially that I load a lot of data but still should work - I need to run the whole script only one maybe two times) But I don't know how many rows the new matrix is going to have. Probably sth around 200, but it changes.
> >
> > I used the cat function to merge the 3d matrices:
> > new_data=[];
> > for i=1:10
> > [data]=function
> > new_data=cat(3, new_data, data)
> > end
> > so far it seems to work (but I test my script loading only small amount of data)
> > the other issue I was not able to resolve was with adding descriptions to the matrix (strings)
> > so I want to have numbers in matrix but in each row I want to add a string indicating e.g. subject ID (need to be a string)
> > e.g. [20 30 1 2 3 44 5 kk-008]
> > are there such combined matrices? I couldn't find anything helpful so far. Maybe someone has an idea?
> >
> > Thanks for all your help :)
>
> again, most CSSMers would use a CELL approach (as shown above) and do the concatenation in one shot after the loop...
>
> us

I will try the CELL option (I need to read more about it first), but coming back to the string issue - the string values need to be added much earlier (I mean the function that generates the data need to give such combine matrix - numeric+string and then the loop I showed here will merge matrices generated by the function). So the strings cannot be added as the last step.
is sth like that even possible?
From: James Tursa on 25 Jun 2010 12:07
"Kara Charaziak" <pink03(a)o2.pl> wrote in message <i02i7g$g56$1(a)fred.mathworks.com>...
>
> I will try the CELL option (I need to read more about it first), but coming back to the string issue - the string values need to be added much earlier (I mean the function that generates the data need to give such combine matrix - numeric+string and then the loop I showed here will merge matrices generated by the function). So the strings cannot be added as the last step.
> is sth like that even possible?

For multiple strings you want to keep track of, each potentially a different length, a CELL approach is usually best. It will definitely be worth your while to learn about cell arrays.

James Tursa
First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4
Prev: Delaunay/interpolation from triangular mesh
Next: Help in some matrix operations