CODEWITHSUNDEEP
arraysmatlabloopsmatrixmultidimensional-array
user5618251
user5618251

Reputation: 361

Create 3D matrix in matlab

I want to plot the temporal evolution of the factor of safety (FS, a quantification for the risk of landslides in a certain area).

This is calculated as follows:

effcohesion=0;
rootcohesion=0;
gammat=12.9E3;
gammaw=9810;
c=0;
deltac=0;
m=0.5;
z=2.5;
phi=16;
slope=rand(20,20)*30       % slope of a gridpoint in area

Strength = c + deltac + (gammat - gammaw.*m).*z.*(cosd(slope).^2);
Stress = gammat.*z.*(sind(slope)).*(cosd(slope));
Part = tand(phi);
FS2 = (Strength./Stress).*(Part)

Now. The value of m (= the height of the water table, which determines the FS) varies throughout the year and is thus not constant. I have a file with precipitation, evaporation, etc. data but to make it not too complicated, I here assume that m is just a function of the day of the year:

mnew=zeros(365,1);
for t=1:365
mnew(t)=(m+t)/150;
end

I now have a dataset with FS for 20x20 points where m =0.5 (=FS2) and a file with the evolution of m during the year (= mnew).

How can I now create a 3D matrix where (1) the spatial variation of FS is stored (so the values of FS over the 20x20 matrix) and (2) the temporal evolution of FS in function of m throughout the year. Eventually, I want a matrix that has both the spatial and temporal evolution of FS in it.

Layer 1 = FS at all 20x20 points on day 1

Layer 2 = FS at all 20x20 points on day 2

etc.

Can someone help me?

Thanks in advance!

Upvotes: 2

Views: 134

Answers (1)

Jeff Irwin
Jeff Irwin

Reputation: 1051

A "3D matrix" would more properly be called a rank 3 array. To do this, just paste your FS2 calculation inside the time loop. Instead of m, use the appropriate mnew to calculate FS2. Then set that layer of FS3 (the rank 3 array) to FS2.

Then, layer 1 (day 1) is given by FS3(:,:,1), layer 2 by FS3(:,:,2), etc.

m0=0.5;

% Sizes of array
n1 = 20;
n2 = 20;
n3 = 365;

FS3 = zeros(n1, n2, n3);

mnew=zeros(n3,1);
for t=1:n3

    mnew(t)=(m0+t)/150;

    effcohesion=0;
    rootcohesion=0;
    gammat=12.9E3;
    gammaw=9810;
    c=0;
    deltac=0;

    m = mnew(t);

    z=2.5;
    phi=16;
    slope=rand(n1,n2)*30;       % slope of a gridpoint in area

    Strength = c + deltac + (gammat - gammaw.*m).*z.*(cosd(slope).^2);
    Stress = gammat.*z.*(sind(slope)).*(cosd(slope));
    Part = tand(phi);
    FS2 = (Strength./Stress).*(Part);

    FS3(:,:,t) = FS2;

end

Upvotes: 2

Related Questions