Cosine similarity between two matrices
127 vues (au cours des 30 derniers jours)
Afficher commentaires plus anciens
Dear all,
I have some vectors 32x1, representing force fields. I use the quiver function to plot and visualize the fields. (see pic attached) Obtaining a quiver showing 4x4 vectors, for a total of 16 vectors.
X = [1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4];
Y = [1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4];
U = W(1:2:31);
V = W(2:2:32);
figure, quiver(X,Y,U',V');
Even if visually they look very similar, I need to calculate a cosine similarity value, between the different vectors.
Checking online I found that this formula:
cosSim = dot(a,b)/(norm(a)*norm(b));
and also the function
D = pdist(X,'cosine').
I wonder if these two functions can be applied to my case and which one is better to obtain just a similarity value for each couple of vector compared.
Thank you!
0 commentaires
Réponse acceptée
Matt J
le 24 Mai 2022
I tend to prefer pdist2,
UV=[U(:),V(:)];
D=1-pdist2(UV,UV,'cosine')
Plus de réponses (0)
Voir également
Catégories
En savoir plus sur Vector Fields 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!