I don't understand why does it give error?

I downloaded the following code from Mathworks site whose URL is :

URL: https://www.mathworks.com/matlabcentral/fileexchange/52949-the-butterfly-particle-swarm-optimization-butterfly-pso-bf-pso-technique?s_tid=FX_rc1_behav

But when I run it, it gives an error. The code is:

%######### Example_Functions.m file ###########
%#### Write the several benchmark functions for BF-PSO here example of sphere is given ###
function fitns_fn=Fitness_fn(x) % Sphere function
fitns_fn=sum(x.^2);
end
%########### Function_limit.m file #######
%#### Give the limits of benchmark functions for BF-PSO ###
function[minm,maxm]=Function_limit(Example_Function) 
    minm=-100; % minimum value of dmension (or variables)
    maxm=100; % maximum value of dmension (or variables)
end
%########### BF-PSO.m file #############
%####### Butterfly-PSO (or BF-PSO)#########
clear all
clc
format short
global bf
global dm
global Example_Function
[minm,maxm]=Function_limit(Example_Function);
ft_fn='Fitness_fn';
%############### Initialization ##############
 flight_max  = 1000; % Max number of flights by BF(or iterations) 
itera_max=flight_max;
bf = 20;    % Size of the butterfly swarm
dm = 30;    % dimensions, 2, 10, 20, 30 ( or problem variables)
c1 =2;  % Acceleration rate C1 
c2 =2;  % Acceleration rate C2 
pg=1;
locations = rand(dm, bf).*(maxm-minm)+minm;
%%%% velocity limits as per variable problems;%%%%%% 
vmax=((maxm-minm)/(rand*maxm));
vmin=-vmax;
velocity = rand(dm, bf).*(vmax-vmin)+vmin;
current_fitness = feval(ft_fn,locations);
local_best_locations  = locations;
mean_best_locations=mean(local_best_locations,2);
local_best_fitness  = current_fitness;
[global_best_fitness,Ind] = min(local_best_fitness);
globl_best_locations = local_best_locations(:,Ind);
     %%%%%%%%%%% Main iterations counter Loop %%%%%%%%%
itera=0;
while (itera < itera_max) % start main iteration loop
      itera = itera+1; p=ones(dm,bf)*(global_best_fitness./(sum(local_best_fitness)));  %%% probability of nectar
        ptt=(1-pt);
s=ones(dm, bf)*exp(-(itera_max-itera)/itera_max); % sensitivity of butterflies
w=(itera_max-itera)/itera_max;
velocity=w.*velocity+ptt.*s.*c1.*rand(dm,bf).*(local_best_locations-locations)+...
pg.*c2.*rand(dm,bf).*((globl_best_locations-locations)*ones(1,bf));   % velocity updation
velocity=min(vmax,max(vmin,velocity));% check velocity limits
locations=locations+rand.*pt.*velocity); 
% location updation
locations=min(maxm, max(minm,locations)); % check location or variable limits
%%%%%%%%%%%%%%%%%% Updating local, mean and global best %%%%%%% 
current_fitness=feval(ft_fn,locations);
Ifit=find(current_fitness<local_best_fitness);
local_best_fitness(:,Ifit)=current_fitness(:,Ifit);
local_best_locations(:,Ifit)=locations(:,Ifit);
 [current_global_best_fitness,Ind]=min(local_best_fitness);
    if current_global_best_fitness < global_best_fitness
       global_best_fitness = current_global_best_fitness;
    end 
globl_best_locations=local_best_locations(:,Ind);
minfit2(itera+1)=global_best_fitness; 
% record min fitness value for plot   
end % end of main iteration loop
min_fit=global_best_fitness     
globl_best_locati=globl_best_locations
bf_mean_fit=mean(minfit2)
bf_stdvn_fit=std(minfit2)

%%%%%% plots %%%%%%%
itera1=0:length(minfit2)-1;
plot(itera1, minfit2,'b')
xlabel('Iterations');
ylabel('Fitness Value');
legend('BF-PSO')

Answer 1

[Alan Stevens]

Put the functions at the end. Tidy up one or two minor errors to get:

%########### BF-PSO.m file #############
%####### Butterfly-PSO (or BF-PSO)#########
clear all
clc
format short
global bf
global dm
global Example_Function
[minm,maxm]=Function_limit(Example_Function);
ft_fn='Fitness_fn';
%############### Initialization ##############
 flight_max  = 1000; % Max number of flights by BF(or iterations) 
itera_max=flight_max;
bf = 20;    % Size of the butterfly swarm
dm = 30;    % dimensions, 2, 10, 20, 30 ( or problem variables)
c1 =2;  % Acceleration rate C1 
c2 =2;  % Acceleration rate C2 
pg=1;
locations = rand(dm, bf).*(maxm-minm)+minm;
%%%% velocity limits as per variable problems;%%%%%% 
vmax=((maxm-minm)/(rand*maxm));
vmin=-vmax;
velocity = rand(dm, bf).*(vmax-vmin)+vmin;
current_fitness = feval(ft_fn,locations);
local_best_locations  = locations;
mean_best_locations=mean(local_best_locations,2);
local_best_fitness  = current_fitness;
[global_best_fitness,Ind] = min(local_best_fitness);
globl_best_locations = local_best_locations(:,Ind);
     %%%%%%%%%%% Main iterations counter Loop %%%%%%%%%
itera=0;

% You need to define pt (arbitrary value here)
pt = 0.1;

while (itera < itera_max) % start main iteration loop
      itera = itera+1; p=ones(dm,bf)*(global_best_fitness./(sum(local_best_fitness)));  %%% probability of nectar
        ptt=(1-pt);
s=ones(dm, bf)*exp(-(itera_max-itera)/itera_max); % sensitivity of butterflies
w=(itera_max-itera)/itera_max;
velocity=w.*velocity+ptt.*s.*c1.*rand(dm,bf).*(local_best_locations-locations)+...
pg.*c2.*rand(dm,bf).*((globl_best_locations-locations).*ones(1,bf));   % velocity updation
velocity=min(vmax,max(vmin,velocity));% check velocity limits
locations=locations+rand.*pt.*velocity; 
% location updation
locations=min(maxm, max(minm,locations)); % check location or variable limits
%%%%%%%%%%%%%%%%%% Updating local, mean and global best %%%%%%% 
current_fitness=feval(ft_fn,locations);
Ifit=find(current_fitness<local_best_fitness);
local_best_fitness(:,Ifit)=current_fitness(:,Ifit);
local_best_locations(:,Ifit)=locations(:,Ifit);
 [current_global_best_fitness,Ind]=min(local_best_fitness);
    if current_global_best_fitness < global_best_fitness
       global_best_fitness = current_global_best_fitness;
    end 
globl_best_locations=local_best_locations(:,Ind);
minfit2(itera+1)=global_best_fitness; 
% record min fitness value for plot   
end % end of main iteration loop
min_fit=global_best_fitness;     
globl_best_locati=globl_best_locations;
bf_mean_fit=mean(minfit2);
bf_stdvn_fit=std(minfit2);

%%%%%% plots %%%%%%%
itera1=0:length(minfit2)-1;
plot(itera1, minfit2,'b')
xlabel('Iterations');
ylabel('Fitness Value');
legend('BF-PSO')

%######### Example_Functions.m file ###########
%#### Write the several benchmark functions for BF-PSO here example of sphere is given ###
function fitns_fn=Fitness_fn(x) % Sphere function
fitns_fn=sum(x.^2);
end

%########### Function_limit.m file #######
%#### Give the limits of benchmark functions for BF-PSO ###
function[minm,maxm]=Function_limit(Example_Function) 
    minm=-100; % minimum value of dmension (or variables)
    maxm=100; % maximum value of dmension (or variables)
end

[Sadiq Akbar]

Thank you dear Alan Stevens for your help. Yest it works now. But when I checked it with the following function listed at the very bottom under listed my function , then it doesn't work:

%########### BF-PSO.m file #############
%####### Butterfly-PSO (or BF-PSO)#########
clear all
clc
format short
global bf
global dm
global Example_Function

u=[1 2 30 40];
[minm,maxm]=Function_limit(Example_Function);
ft_fn=@(b)karimVectorized1(b,u);%'Fitness_fn';
%############### Initialization ##############
 flight_max  = 1000; % Max number of flights by BF(or iterations) 
itera_max=flight_max;
bf = 20;    % Size of the butterfly swarm
dm = 4;%30;    % dimensions, 2, 10, 20, 30 ( or problem variables)
c1 =2;  % Acceleration rate C1 
c2 =2;  % Acceleration rate C2 
pg=1;
locations = rand(dm, bf).*(maxm-minm)+minm;
%%%% velocity limits as per variable problems;%%%%%% 
vmax=((maxm-minm)/(rand*maxm));
vmin=-vmax;
velocity = rand(dm, bf).*(vmax-vmin)+vmin;
current_fitness = feval(ft_fn,locations);
local_best_locations  = locations;
mean_best_locations=mean(local_best_locations,2);
local_best_fitness  = current_fitness;
[global_best_fitness,Ind] = min(local_best_fitness);
globl_best_locations = local_best_locations(:,Ind);
     %%%%%%%%%%% Main iterations counter Loop %%%%%%%%%
itera=0;

% You need to define pt (arbitrary value here)
pt = 0.1;

while (itera < itera_max) % start main iteration loop
      itera = itera+1; p=ones(dm,bf)*(global_best_fitness./(sum(local_best_fitness)));  %%% probability of nectar
        ptt=(1-pt);
s=ones(dm, bf)*exp(-(itera_max-itera)/itera_max); % sensitivity of butterflies
w=(itera_max-itera)/itera_max;
velocity=w.*velocity+ptt.*s.*c1.*rand(dm,bf).*(local_best_locations-locations)+...
pg.*c2.*rand(dm,bf).*((globl_best_locations-locations).*ones(1,bf));   % velocity updation
velocity=min(vmax,max(vmin,velocity));% check velocity limits
locations=locations+rand.*pt.*velocity; 
% location updation
locations=min(maxm, max(minm,locations)); % check location or variable limits
%%%%%%%%%%%%%%%%%% Updating local, mean and global best %%%%%%% 
current_fitness=feval(ft_fn,locations);
Ifit=find(current_fitness<local_best_fitness);
local_best_fitness(:,Ifit)=current_fitness(:,Ifit);
local_best_locations(:,Ifit)=locations(:,Ifit);
 [current_global_best_fitness,Ind]=min(local_best_fitness);
    if current_global_best_fitness < global_best_fitness
       global_best_fitness = current_global_best_fitness;
    end 
globl_best_locations=local_best_locations(:,Ind);
minfit2(itera+1)=global_best_fitness; 
% record min fitness value for plot   
end % end of main iteration loop
min_fit=global_best_fitness;     
globl_best_locati=globl_best_locations;
bf_mean_fit=mean(minfit2);
bf_stdvn_fit=std(minfit2);

%%%%%% plots %%%%%%%
itera1=0:length(minfit2)-1;
plot(itera1, minfit2,'b')
xlabel('Iterations');
ylabel('Fitness Value');
legend('BF-PSO')

%######### Example_Functions.m file ###########
%#### Write the several benchmark functions for BF-PSO here example of sphere is given ###
function fitns_fn=Fitness_fn(x) % Sphere function
fitns_fn=sum(x.^2);
end

%########### Function_limit.m file #######
%#### Give the limits of benchmark functions for BF-PSO ###
function[minm,maxm]=Function_limit(Example_Function) 
    minm=-100; % minimum value of dmension (or variables)
    maxm=100; % maximum value of dmension (or variables)
end


%%%%%%%%%%% My Function %%%%%%%%%%%%%%%
function e=karimVectorized1(b,u)
u=u';
b=b';
[R,~] = size(b);
P = R/2;
M = 2*R;
%%%%%%%%%%%%%%%%%%%%%%%%
% Swapping vector b 
%%%%%%%%%%%%%%%%%%%%%%%%
[~, ix] = sort(u); % u is my desired vector
[~, ix1(ix)] = sort(b); 
b = b(ix1);
%%%%%%%%%%%%%%%
% calculate xo
%%%%%%%%%%%%%%%
u1 = u(    1:P );
u2 = u( P+(1:P));
k = 1:M;

xo = sum( u1.*exp( -1i.*pi*cosd(u2).*(k-1) ) , 1);
% xo = awgn(xo,Noise);% add Noise

%%%%%%%%%%%%%%%
% Calculate xe
%%%%%%%%%%%%%%%
b1 = b(    1:P );
b2 = b( P+(1:P));

xe = sum( b1.*exp( -1i.*pi*cosd(b2).*(k-1) ) , 1);

%%%%%%%%%%%%
% MSE
%%%%%%%%%%%%
%e = mean( abs( xo(1,:) - xe(1,:) ) .^2 );% Both are correct
e = mean(abs(xo-xe).^2,2);
end

It gives the following error:

Unable to perform assignment because the indices on the left side are not compatible with

the size of the right side.

Error in BFPSO_CorrectedByMathworks>karimVectorized1 (line 97)

[~, ix1(ix)] = sort(b);

Error in BFPSO_CorrectedByMathworks>@(b)karimVectorized1(b,u) (line 12)

ft_fn=@(b)karimVectorized1(b,u);%'Fitness_fn';

Error in BFPSO_CorrectedByMathworks (line 26)

current_fitness = feval(ft_fn,locations);

[Sadiq Akbar]

@Alan Stevens please help me

[Alan Stevens]

I notice you haven't changed the arbitrary value of p that I used. You need to insert the value appropriate for your problem. I don't know if this will solve the issue or not.

If not, please explicitly note the lines that you have changed from the program I got working.

[Sadiq Akbar]

@Alan Steven

Thank you very much dear Alan Stevens for your kind response. I am giving both the codes in the attachment. One is that corrected by you called "BFPSO_AlanStevens.m" and one is that changed by me called "BFPSO_mine.m". As you can see "BFPSO_mine" is also the same as is yours. But this shows the changes where I have made in yours. It works from the start lines but gives the error where it calls my function. And my function is given at the very bottom of the whole code. You can locate it very easily now. But why it is giving an error, that I don't undesratnd. Further, it also includes the value of Pt you defined. But the reason of error is not understood by me. The only difference in the two functions is that my function has 4 variables and has lower bound as [0 0 0 0] and upper bound as [10 10 90 90] but the original function has 30 variables and all have lower bound as -100 and upper bound as 100.

Regards,

[Alan Stevens]

In your function b1 and b2 have different sizes from k, hence you can't do element by element multiplication as required in xe. I don't know enough about exactly what your program is modelling to help any further,

%%%%%%%%%%% My Function %%%%%%%%%%%%%%% 
function e=karimVectorized1(b,u) %My function------------(8)

u=u';
b=b';
[R,~] = size(b);
P = R/2;
M = 2*R;
%%%%%%%%%%%%%%%%%%%%%%%%
% Swapping vector b 
%%%%%%%%%%%%%%%%%%%%%%%%
[~, ix] = sort(u); % u is my desired vector
[~, ix1(ix,:)] = sort(b); %%%%%%%%%%%%%%%%%%%%%%%%%%%
b = b(ix1);

%%%%%%%%%%%%%%%
% calculate xo
%%%%%%%%%%%%%%%
u1 = u(    1:P );
u2 = u( P+(1:P));
k = 1:M;

xo = sum( u1.*exp( -1i.*pi*cosd(u2).*(k-1) ) , 1);
% xo = awgn(xo,Noise);% add Noise

%%%%%%%%%%%%%%%
% Calculate xe
%%%%%%%%%%%%%%%
b1 = b(    1:P );
b2 = b( P+(1:P));


% b1 and b2 have size 1x2, but k has size 1x8 hence the bracketed term 
% in the following line cannot be executed.  
xe = sum( b1.*exp( -1i.*pi*cosd(b2).*(k-1) ) , 1);


%%%%%%%%%%%%
% MSE
%%%%%%%%%%%%
%e = mean( abs( xo(1,:) - xe(1,:) ) .^2 );% Both are correct

e = mean(abs(xo-xe).^2,2);

end