Reduce 3D array by removing NaN elements

Hi !
I ask you for help in Matlab in order to reduce a 3D array. In column there is three numbers (X,Y,Z coordinates) and each row corresponds to a different point. But these points have to respect a certain condition, otherwise three NaN are stored instead of the coordinates. The third dimension correspond to the time evolution of the coordinates stored in the 2D array.
For exemple:
A(:,:,1) =
1 2 3
NaN NaN NaN
NaN NaN NaN
A(:,:,2) =
NaN NaN NaN
1 2 3
NaN NaN NaN
A(:,:,3) =
NaN NaN NaN
NaN NaN NaN
NaN NaN NaN
If I try:
A(any(any(isnan(A),3),2),:,:) = [];
I get:
Empty array: 0-by-3-by-3
Or I would have:
A(:,:,1) =
1 2 3
NaN NaN NaN
A(:,:,2) =
NaN NaN NaN
1 2 3
A(:,:,3) =
NaN NaN NaN
NaN NaN NaN
The array is reduced but the shape is the same and I don't lose informations.
Thank your for your help and have a nice day !

2 comentarios

the cyclist
the cyclist el 25 de Ag. de 2014
What exactly do you want to get rid of? Any row in which all time slices are NaN?
No, in fact my 3D array is very huge (50,000 by 3 by 40). I have to display this array on screen, so there is 2,000,000 lines ! It's very huge and it takes a long time to display so many lines. I just want to reduce this 3D array by removing as many "NaN-lines" as I can without erease "coordinates-lines", unlike to the command:
A(any(any(isnan(A),3),2),:,:) = [];

Iniciar sesión para comentar.

 Respuesta aceptada

the cyclist
the cyclist el 25 de Ag. de 2014
Editada: the cyclist el 25 de Ag. de 2014
Maybe you meant all() instead of any()?
idxToRemove = all(all(isnan(A),3),2);
A(idxToRemove,:,:) = []

Más respuestas (1)

Anthony
Anthony el 25 de Ag. de 2014

0 votos

It works perfectly ! Thank you very much ! Could you explain to me how this command works please ? I would understand and not just apply without thinking.

Etiquetas

Preguntada:

el 25 de Ag. de 2014

Respondida:

el 25 de Ag. de 2014

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by