How to find the indices of non-zero elements in a matrix
66 vues (au cours des 30 derniers jours)
Afficher commentaires plus anciens
Hello,
I have an Na-by-Nt matrix which is sparse, i.e.: most elements are zeros. I want to find the indices of the non-zeros elements in the form of (i,j) where i is the row and j is the column.
I have then these two Na-by-1 vector a and Nt-by-1 vector tau. i corresponds to the ith element in a and j the jth element in tau, and I want to find them as well.
How can I do the above in an efficient way without the need for for loops?
Thanks
0 commentaires
Réponses (1)
Cedric
le 27 Juin 2014
Modifié(e) : Cedric
le 27 Juin 2014
If it is for storing only non-zero elements, MATLAB supports SPARSE matrices. If you need to get row/column indices of non-zero elements of A:
[rId, cId] = find( A ) ;
and if you need values as well:
[rId, cId, val] = find( A ) ;
If you just need values, you can get them simply with
vals = A(A ~= 0) ;
11 commentaires
Cedric
le 30 Juin 2014
Modifié(e) : Cedric
le 30 Juin 2014
I meant a simple/minimal numeric example showing which triplets you need to ID in which data structure, because I am not sure that anybody will have time to understand your code/algorithm.
If it is working but too slow, the first thing that comes to my mind is to replace the last double FOR loop with
A = complex( zeros( numel(s), Np * Ntau )) ;
colId = 0 ;
for pp=1:Np
theta_p=fc.*kron(1+Doppler_Tentative_Set(pp),ones(N))-fr;
Gamma_p=exp(1i*pi*T.*theta_p).*sinc(T.*theta_p);
for tt=1:Ntau
Lambda_p_diag = exp(-1i*2*pi.*(f0+(0:N-1)/T).*Tau_Tentative_Set(tt));
colId = colId + 1 ;
A(:,colId) = bsxfun( @mtimes, Gamma_p, Lambda_p_diag ) * s ;
end
end
Voir également
Catégories
En savoir plus sur Creating and Concatenating Matrices dans Help Center et File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!