Gridded Interpolation in a certain volume and leave outside with NaN values

the following scrpt creates this plot:

load('var.mat')
scatter3(xq(:),yq(:),zq(:),2,'k','filled'); hold on
scatter3(x,y,z,10,'r','filled')

From the data, you can see that xq,yq,zq and cq are a gridded data.

I want to use griddedInterpolant function to "fill" the cq(which contains NaN) values inside to x,y,z,c positions with an interpolation of c info. and remain all the points "outside" the red points as NaN.

Answer 1

[KSSV]

load('var.mat') ; 

scatter3(xq(:),yq(:),zq(:),2,'k','filled'); hold on
scatter3(x,y,z,10,'r','filled')

F = scatteredInterpolant(x,y,z,c,'nearest') ; 
cq0 = cq ; 
for i = 1:583
    i
    cq(:,:,i) = F(xq(:,:,i),yq(:,:,i),zq(:,:,i)) ;
end

[Philippe Corner]

Thank you very much. Yes, your suggestion interpolates to the black points perfectly. But the points "outside" a boundary that the red points may create are extrapolated, and not with NaN, is it possible to avoid extrapolation with scatteredInterpolant? do you have an idea about how to change them to NaN?

[KSSV]

Read the document......you have an option to play around with that..you need to use Extrapolation method to none.

[Philippe Corner]

Yes, you were right. It worked now. I appreciate a lot your help