What is the modification needed in the below code

Are the below code same? I am getting different results.

%First
scale0 = 1500;
Nai = linspace(1e13,1e19,scale0);   %𝑁𝑎:1×1013−1×1019𝑐m−3
Ndi = linspace(1e13,1e19,scale0);   %𝑁𝑑:1×1013−1×1019𝑐m−3
Na = ones(scale0,1)*Nai;
Nd = Ndi.'*ones(1,scale0);

y = 0;
z = 0;
%Computing Fermi Energy Level
Ef = ones(scale0,1)*0;
while y < 1500
    y = y + 1;
    while z < 1500
        z = z + 1;
        eq1 = @(Ef) ((Nc)*exp(-(Ec-Ef)/(kbT))) + ((Na(y,z))/(1 + 4*exp(-(Ef-Ea)/(kbT)))) - ((Nv)*exp(-(Ef-Ev)/(kbT)) + ((Nd(y,z))/(1 + 2*exp(-(Ed-Ef)/(kbT)))));
        x1 = [0 10];
        Ef(y,z) = fzero(eq1,x1);
    end
end


%Second
scale0 = 1500;
Nai = linspace(1e13,1e19,scale0);   %𝑁𝑎:1×1013−1×1019𝑐m−3
Ndi = linspace(1e13,1e19,scale0);   %𝑁𝑑:1×1013−1×1019𝑐m−3
Na = ones(scale0,1)*Nai;
Nd = Ndi.'*ones(1,scale0);

%Computing Fermi Energy Level
Ef = ones(scale0,1)*0;
for y=1:prod(size(Nai))
    for z=1:prod(size(Ndi))
        eq1 = @(Ef) ((Nc)*exp(-(Ec-Ef)/(kbT))) + ((Na(y,z))/(1 + 4*exp(-(Ef-Ea)/(kbT)))) - ((Nv)*exp(-(Ef-Ev)/(kbT)) + ((Nd(y,z))/(1 + 2*exp(-(Ed-Ef)/(kbT)))));
        x1 = [0 10];
        Ef(y,z) = fzero(eq1,x1);
    end
end

[Rena Berman]

(Answers Dev) Restored edit

Answer 1

[Walter Roberson]

Look at the first set of code. You initialize z = 0 outside both loops. In the inner loop you add 1 to each z, ending when z >= 1500. Now consider the second iteration of the while y loop, when y increments from 1 to 2. z is still 1500 from the previous while z<1500 loop, because you never reset z to 0. So for the rest of the y iterations, you do no z work.

Side note:

prod(size(Nai))

can be written more efficiently as

numel(Nai)

[Amy Topaz]

Thank you so much.

How can I implement the while loop to produce the same result as the for loop?

[Walter Roberson]

Initialize z within the first while loop.

[Amy Topaz]

thank you