need help on genetic algorithm

hi I have this code I run it on matlab command its ok working I try to make it as a fitness function for GA but its not working at all

%function MSE= GA(g_pulse_max,x1train_max)

Xrange=-3:0.1:2.4;

mu=-0.5;

sigma= 0.17;

g_pulse= 1/sqrt(2*pi*sigma^2) * exp( - ( Xrange - mu).^2/(2*sigma^2));

g_pulse_max=g_pulse/max(g_pulse);

plot(g_pulse_max);

hold on 

x1train_max = x1train/max(x1train);

plot(x1train_max);

hold on

can anybody help me to solve it?

Answer 1

[Walter Roberson]

What error shows up?

You would certainly not want to be plotting in a fitness function.

Your commented out header for the code suggests that you expect two variables to be passed in. The fitness function for ga() will pass in only a single variable which will be a row vector whose length is nvar the number of variables you declared were present.

Your commented out header for the code suggests that you expect to calculate a variable named MSE and return it. Your code does not calculate MSE.

Answer 2

[Asmaa Mohammed]

thank you so much for reply.. what I want from the code is to enter normal gaussian pulse with normalized data sequence I want for the GA to shape the data sequence like gaussian pulse in other words I want to remove the noise or the lower impulse response from the data sequence how can write the fitness function to achieve that? after that I get the output from the GA and compute the mean square error. I will be so thankful if you help me to achieve that.

[Asmaa Mohammed]

like this photo.. the green curve represent the gaussian pulse the red curve is for distorted signal.. I want to redistribute the impulse response for the red curve to be fitted inside the gussian pulse by using GA