I wanna do the plot exactly the same as the attached picture, my Y axis should start from 0.0001. and i need to draw this red dashed line from the end of plot point to x axis

clear all;
load 'Theta_x.txt';               %correction factor 1-4
X=Theta_x;
load 'POE_Y.txt';                 %probability of exceedance of the factor HEC-18 scour
Y=POE_Y;
semilogy(X,Y);
title('Correction Factor And Probability of Exceedance');
xlabel('Correction Factor Θz');
ylabel('Probability of Exceedance PoE');
Theta_x = readmatrix('Theta_x.txt');
POE_Y = readmatrix('POE_Y.txt');
X=Theta_x;
Y=POE_Y;
plot(X,Y)
grid
set(gca, 'YScale','log')
title('Correction Factor And Probability of Exceedance');
xlabel('Correction Factor Θz');
ylabel('Probability of Exceedance PoE');

Answer 1

[Voss]

How about this?

Theta_x = readmatrix('Theta_x.txt');
POE_Y = readmatrix('POE_Y.txt');
plot(Theta_x,POE_Y)

hold on
plot([max(Theta_x) 3],[min(POE_Y) 1e-4],'--r')

grid
set(gca, 'YScale','log')
title('Correction Factor And Probability of Exceedance');
xlabel('Correction Factor Θz');
ylabel('Probability of Exceedance PoE');

xlim([0 4])
ylim([1e-4 1])

[Khaled aL Qattawi]

that is perfect. thanks.

i have one question, i see you here you defind the where the red line hit the x axis at 3. is there anyway we can make the matlab forcast this line??

[Voss]

Sure:

Theta_x = readmatrix('Theta_x.txt');
POE_Y = readmatrix('POE_Y.txt');
plot(Theta_x,POE_Y)

% fit a line to the (log) data:
% use log10(POE_Y) as the independent variable and 
% Theta_x as the dependent variable ...
p = polyfit(log10(POE_Y),Theta_x,1);
% ... because we want to know the x value where 
% the line hits y == -4
x_axis_crossing = polyval(p,log10(1e-4))

x_axis_crossing = 2.9909

% (turns out to be very close to 3)

hold on
plot([max(Theta_x) x_axis_crossing],[min(POE_Y) 1e-4],'--r')

grid
set(gca, 'YScale','log')
title('Correction Factor And Probability of Exceedance');
xlabel('Correction Factor Θz');
ylabel('Probability of Exceedance PoE');

xlim([0 4])
ylim([1e-4 1])

[Khaled aL Qattawi]

thank you , thank you, thank you, i am really glad.

[Voss]

For your reference, this is also showing the line that was fit to the data and used to extrapolate to the x-axis:

Theta_x = readmatrix('Theta_x.txt');
POE_Y = readmatrix('POE_Y.txt');
plot(Theta_x,POE_Y)

p = polyfit(log10(POE_Y),Theta_x,1);
x_axis_crossing = polyval(p,log10(1e-4))

x_axis_crossing = 2.9909

hold on
plot([max(Theta_x) x_axis_crossing],[min(POE_Y) 1e-4],'--r')

% plotting the fitted line as well:
plot(polyval(p,log10([1 1e-4])),[1 1e-4],':g','LineWidth',3)

grid
set(gca, 'YScale','log')
title('Correction Factor And Probability of Exceedance');
xlabel('Correction Factor Θz');
ylabel('Probability of Exceedance PoE');

xlim([0 4])
ylim([1e-4 1])

[Voss]

You're welcome! Any other questions, please let me know.

[Khaled aL Qattawi]

that is perfect, it will help me a lot too. thank you.

[Voss]

You're welcome!

[Khaled aL Qattawi]

thanks, but i have another question, in the top we used straight line or linear process to forcast the rest of the curve, but some times this doesn't work correctly. please the attached picture for different set of data show something doesn't fit good, is there other comand this this type of fit curves.