Logarithmic scale problem in 2D graph

Hello! I am trying to program equation (2) and make graph number 1, specifically the curve that Masetti [21] says, but when I do it I have an error where the graph starts, could someone please help me :(

Graph 1

Mine :(

My code:

u1=68.5;

umax=1414;

u2=56.1;

Cr=9.2e16;

Cs=3.41e20;

alfa=0.711;

beta=1.98;

N1=1e16;

N2=1e16;

N3=1e21;

index=0;

k = N1:N2:N3;

for Ne = 1:length(k)

index = index + 1 ;

u(index) = u1 + ((umax-u1)/(1+(k(Ne)/Cr)^alfa)) - ((u2)/(1+(Cs/k(Ne))^beta));

f(index) = k(Ne);

x(index) = f(index) ;

y(index) = u(index);

end

semilogx(x,y)

yticklabels({'0','100','200','300','400','500'})

[Karim]

When I try it (see below) I obtain the same curve as you have. The values and the equation are the same as from the snapshot of the paper. Hence, I would say that your curve, also at the start, is indeed correct.

If you want to obtain the same curve as 'graph1' you will need another set of parameters. Are you certain that those parameters corespond to that graph?

EDIT: disabled the ytick label text, as this could lead to wrong interpretation of the data

mu1 = 68.5;              
mumax = 1414;              
mu2 = 56.1;                 
Cr = 9.2e16;                
Cs = 3.41e20;               
alfa = 0.711;               
beta = 1.98;                

Ne = logspace(16,21,1000);
mu = mu1 + ( (mumax-mu1) ./ (1+(Ne./Cr).^alfa) ) - ( mu2 ./ (1+(Cs./Ne).^beta) );

figure
semilogx(Ne,mu)
%yticklabels({'0','100','200','300','400','500'})
grid on
xlabel('$N_{e}$','interpreter','latex','FontSize',20)
ylabel('$\mu$','interpreter','latex','FontSize',20)

Answer 1

[Star Strider]

I don’t see anything wrong with the code (however I rewrote the μ assignment to make it a bit easier for me to understand in the context of the equation provided).

I suspect there’s something wrong with the y-axis scaling (as expressed in the yticklabels call), however not knowing what that scaling multiplication is, I can’t determine if there’s a problem.

Plotting it without the scaling —

u1=68.5;                  
umax=1414;                
u2=56.1;                 
Cr=9.2e16;                
Cs=3.41e20;                         
alfa=0.711;               
beta=1.98;                
N1=1e16;
N2=1e16;
N3=1e21;
Ne = N1:N2:N3;
for k = 1:length(Ne)
    u(k) = u1 + ((umax-u1)/(1+(Ne(k)/Cr)^alfa)) - (u2/(1+(Cs/Ne(k))^beta));
end 
semilogx(Ne,u)
xlabel('Ne')
ylabel('\mu')

% yticklabels({'0','100','200','300','400','500'})

.

[Markus M.]

Did try it too and came to the same conclusion as @Star Strider. I don't see a reason to change the y-tick labels to "wrong" numbers. Maybe the plot from the paper did use other parameters? Or even another unit for ?

[Yordani]

I already found the mistake! Thanks for your contribution!

[Star Strider]

My pleasure!

Just out of curiosity, what was the mistake? We’d all like to know!