repeat a loop for 10 times

I try to create a loop for which I want to be repeated 10 times. 10 times are enough iterations in order to have at last a constant number.

The loop I create is the following.

I suppose for L_repeat a number (-9999) and after 10 iterations I expect to have the solution in the equation of L_repeat. At the end of each iteration the solution of the equation of L_repeat will be the numbers used in the beginning of the next iteration.

for iterations=10

    L_repeat=[-9999; -9999; -9999; -9999; -9999; -9999; -9999; -9999]

    esh_repeat=zm./L_repeat

    if Ri_zm(:,1)<0
        phi_m_repeat=(1-19.3.*esh_repeat).^(-1/4)
        phi_H_repeat=0.95.*((1-11.6.*esh_repeat).^(-1/2))
    else if Ri_zm(:,1)>0 & Ri_zm(:,1)<=0.2
            phi_m_repeat=1+6.*esh_repeat
            phi_H_repeat=0.95+(7.8.*esh_repeat)
        end
    end

    % neutral
    % φm(ς)=φΗ(ς)= φΕ(ς)=1
    if Ri_zm(:,1)<=0.01 & Ri_zm(:,1)>=-0.01
        u_star_repeat=0.40./p_u(:,1)
        theta_star_repeat=0.40./p_theta(:,1)
    % unstable και stable (Ri<=0.2)
    else if Ri_zm(:,1)<-0.01 | Ri_zm(:,1)>0.01 & Ri_zm(:,1)<=0.2
    u_star_repeat=0.40./(p_u(:,1).*phi_m_repeat(:,1))
    theta_star_repeat=0.40./(p_theta(:,1).*phi_H_repeat(:,1))
        end
    end

    L_repeat=((u_star_repeat.^2).*theta_horizontal_mean)./(0.40.*9.80665.*theta_star_repeat)

end

[Matt Kindig]

I'm confused. Won't your last line (L_repeat=...) be overwritten at the beginning of the next iteration?

[christiana]

Exactly. This new loop is very close to that I want. But somewhere I am wrong.

for j=1:9

    j=j+1

    L_repeat(:,1)=[-9999; -9999; -9999; -9999; -9999; -9999; -9999; -9999]

    esh_repeat=zm./L_repeat(:,j)

    if Ri_zm(:,1)<0
        phi_m_repeat=(1-19.3.*esh_repeat).^(-1/4)
        phi_H_repeat=0.95.*((1-11.6.*esh_repeat).^(-1/2))
    else if Ri_zm(:,1)>0 & Ri_zm(:,1)<=0.2
            phi_m_repeat=1+6.*esh_repeat
            phi_H_repeat=0.95+(7.8.*esh_repeat)
        end
    end

    % neutral
    % φm(ς)=φΗ(ς)= φΕ(ς)=1
    if Ri_zm(:,1)<=0.01 & Ri_zm(:,1)>=-0.01
        u_star_repeat=0.40./p_u(:,1)
        theta_star_repeat=0.40./p_theta(:,1)
    % unstable και stable (Ri<=0.2)
    else if Ri_zm(:,1)<-0.01 | Ri_zm(:,1)>0.01 & Ri_zm(:,1)<=0.2
    u_star_repeat=0.40./(p_u(:,1).*phi_m_repeat(:,1))
    theta_star_repeat=0.40./(p_theta(:,1).*phi_H_repeat(:,1))
        end
    end

    L_repeat(:,j)=((u_star_repeat.^2).*theta_horizontal_mean)./(0.40.*9.80665.*theta_star_repeat)

end

[christiana]

I think this loop works. This is the answer.

for j=1:10

    L_repeat(:,j)=((u_star_repeat.^2).*theta_horizontal_mean)./(0.40.*9.80665.*theta_star_repeat)

    L_repeat(:,1)=[-9999; -9999; -9999; -9999; -9999; -9999; -9999; -9999]

    esh_repeat=zm./L_repeat(:,j)

    if Ri_zm(:,1)<0
        phi_m_repeat=(1-19.3.*esh_repeat).^(-1/4)
        phi_H_repeat=0.95.*((1-11.6.*esh_repeat).^(-1/2))
    else if Ri_zm(:,1)>0 & Ri_zm(:,1)<=0.2
            phi_m_repeat=1+6.*esh_repeat
            phi_H_repeat=0.95+(7.8.*esh_repeat)
        end
    end

    % neutral
    % φm(ς)=φΗ(ς)= φΕ(ς)=1
    if Ri_zm(:,1)<=0.01 & Ri_zm(:,1)>=-0.01
        u_star_repeat=0.40./p_u(:,1)
        theta_star_repeat=0.40./p_theta(:,1)
    % unstable και stable (Ri<=0.2)
    else if Ri_zm(:,1)<-0.01 | Ri_zm(:,1)>0.01 & Ri_zm(:,1)<=0.2
    u_star_repeat=0.40./(p_u(:,1).*phi_m_repeat(:,1))
    theta_star_repeat=0.40./(p_theta(:,1).*phi_H_repeat(:,1))
        end
    end

end

Answer 1

[Cedric]

Your for loop statement is not built correctly:

 for iterations=10

will just give iterations the value 10 and will execute what follows only once. You have to have a vector of values on the right hand side, that defines all the values that the loop variable has to take iteratively. For example:

 for iterations = 1 : 10
    ...
 end

I have a doubt about the rest of the code as well. My advice would be to build it progressively, displaying what happens. For example, make a first step:

 for ii = 1 : 10
    ii
 end

and observe what happens. Then add a first, small computation in the loop, and check again that it is working, until you have implemented the full code in a well controlled way.

[christiana]

The folowing loop works better. I have a table L_repeat(i,j). At the beginning I give values for the 1st column L_repeat(:,1). At the end of each loop I expect to add one more column in the "L_repeat" table and use this new column for calculations in this loop. After 10 iterations I expext to have the final results in the column L_repeat(:,10).

Any mistake in this new loop?

for j=1:9

    j=j+1

    L_repeat(:,1)=[-9999; -9999; -9999; -9999; -9999; -9999; -9999; -9999]

    esh_repeat=zm./L_repeat(:,j)

    if Ri_zm(:,1)<0
        phi_m_repeat=(1-19.3.*esh_repeat).^(-1/4)
        phi_H_repeat=0.95.*((1-11.6.*esh_repeat).^(-1/2))
    else if Ri_zm(:,1)>0 & Ri_zm(:,1)<=0.2
            phi_m_repeat=1+6.*esh_repeat
            phi_H_repeat=0.95+(7.8.*esh_repeat)
        end
    end

    % neutral
    % φm(ς)=φΗ(ς)= φΕ(ς)=1
    if Ri_zm(:,1)<=0.01 & Ri_zm(:,1)>=-0.01
        u_star_repeat=0.40./p_u(:,1)
        theta_star_repeat=0.40./p_theta(:,1)
    % unstable και stable (Ri<=0.2)
    else if Ri_zm(:,1)<-0.01 | Ri_zm(:,1)>0.01 & Ri_zm(:,1)<=0.2
    u_star_repeat=0.40./(p_u(:,1).*phi_m_repeat(:,1))
    theta_star_repeat=0.40./(p_theta(:,1).*phi_H_repeat(:,1))
        end
    end

    L_repeat(:,j)=((u_star_repeat.^2).*theta_horizontal_mean)./(0.40.*9.80665.*theta_star_repeat)

end

[Cedric]

If you just test

 for ii = 1 : 10
    ii
 end

You will see that the for loop iterates ii already. No need to do it yourself (with j=j+1). For the rest, I stick to my comment, that you should start with just a simple loop, and then progressively implement your computations, checking that each step produces the correct output.

Just a few additional remarks: for a purpose of efficiency, you could preallocate memory for L_repeat, before starting the loop. Also, you don't need to set L_repeat(:,1) at each step of your iteration.

 L_repeat = zeros(8,10) ;
 L_repeat(:,1) = -9999 ;
 for jj = 2 : 10
    L_repeat(:,jj) = ... something computed from L_repeat(:,jj-1) ?
 end