I am newbie in the matlab, may some experts tell me what are the meaning of codes or logic in this file?

Hello, I am a new user to MatLab, but I actually don't actually know what i should type in the matlab program to let me run my desired program.

Actually, I am doing my final year project with the mitigation and the simulation of the bus stop. In order to simulate the real scenario, I use a distribution and a program flow to simulate the bus congestion.

However, as I am new to Matlab, my program files are given by my supervisor and I don't know what is the logic flow or meaning of the codes in the files.

May some experts tell me the flow of the program codes and help me? Thank you very much.

The following are the codes.

Files: Sim_FIFO_MDc.m

% code

global size;
global Avg_Wait;
global Avg_Wait_Q;    % Average wait time in queue per bus
global Serv_Rate;
global Arr_Rate;

global U0;  %parameters same as Sim.m
global j0;
global IntArr;
global Arr;
global Serv;
global Wait;
global Wait_Q;
global Wait_S;
global Pos;
global c;
global Cs;

% Generate inter-arrival time vector
IntArr(1 : size) = -1 / Arr_Rate * log(U0(j0 : j0 + size - 1));    
 % Inter-arrival time follows an exponential distribution (i.e., bus arrivals are Poisson)
j0 = j0 + size;

% Generate service time vector
Serv(1 : size) = (1 / Serv_Rate) * ones(size, 1);   % Service time is deterministic

% Serv(1 : size) = gamrnd(1 / Cs ^ 2, Cs ^ 2 * Serv_Rate, size, 1); 
 % Service time follows gamma distribution, where mean is 1, and coefficient of variation is Cs
Arr(1 : size) = cumsum(IntArr);    % Calculate arrival times
Wait_Q(1) = 0;
Wait_S(1) = 0;
Pos(1) = 1;

for i = 2 : size    % iterate for all the buses
    if Pos(i - 1) == c
        Wait_Q(i) = max(0, Wait_Q(i - 1) + Serv(i - 1) + Wait_S(i - 1) - IntArr(i));
    else
        Wait_Q(i) = max(0, Wait_Q(i - 1) - IntArr(i));
    end

      if Wait_Q(i - 1) + Serv(i - 1) + Wait_S(i - 1) < IntArr(i) + Wait_Q(i) + 0.0001
          Pos(i) = 1;
      else
          Pos(i) = Pos(i - 1) + 1;
      end

      Wait_S(i) = max(0, Wait_Q(i - 1) + Serv(i - 1) + Wait_S(i - 1) - Wait_Q(i) - Serv(i) - IntArr(i));
  end

Wait = Wait_Q + Wait_S;

Avg_Wait = mean(Wait);
Avg_Wait_Q = mean(Wait_Q);

end

% code

Files: Sim.m

% code

T_start = clock;  % see if the program ends or not, ctrl+c can stop the program

global size;      % number of buses in each round go into the bus stop
size = 200000;    % set it into a high size because to reduce the fluctuation error
global Avg_Wait;    % Average wait time per bus
global Serv_Rate;  % Service rate (inverse of average service time per bus)
Serv_Rate = 1;    %set service rate to one , don't care the unit at all.
global Arr_Rate;  % Arrival rate of buses (i.e., bus flow) (eg, 3 buses per time unit(serv_rate))

global U0;      %used for indexes
global j0;
      %all the following global variables is having the size of "size".
global IntArr;    % Vector for inter-arrival times (headway) 
global Arr;    % Vector for arrival times
global Serv;    % Vector for service times
global Wait_Q;    % Vector for bus wait times in queue
global Wait_S;    % Vector for bus wait times in berths
global Wait;    % Vector for bus wait times (sum of Q+S)
global Pos;    % Vector for the berth numbers where buses dwell
global c;    % Berth number (0,1,2) max=2 berths
global Cs;    % Coefficient of variation in service time

M = 20000000;    %random number generator
U0 = rand(M, 1);  % Random seed (a vector of random numbers of size M, each is uniformly distributed between 0 and 1)
j0 = 1;        %random take one number from 0 to 1

IntArr = zeros(size - 1, 1);
Arr = zeros(size, 1);
Serv = zeros(size, 1);
Wait = zeros(size, 1);
Wait_Q = zeros(size, 1);
Wait_S = zeros(size, 1);
Pos = zeros(size, 1);

Cs = 0.5;  % coefficient of variation in service time
c = 2;   % number of berths

%plotting the figure start from 2 because first one always zero.
for i0 = 2:20    % Need 20 data points on the curve
  % Arr_Rate = 0.15, 0.25, ..., 1.95, subject to adjustment
    Arr_Rate = 0.1 * (i0-1) + 0.05;  
    Sim_FIFO_MDc;     %deterministic function service time
    beta(i0) = Arr_Rate / c;   % beta = bus flow per berth
    W(i0) = Avg_Wait;     % Average wait time per bus

      if j0 > M - 200 * size   %reset the U0 and j0 if data is not enough
          U0 = rand(M, 1);
          j0 = 1;
      end
  end

% First data point
beta(1) = 0;
W(1) = 0;

plot(beta, W);

T_end = clock;
T_dur = T_end - T_start;
fprintf('\nThe simulation takes %4.2f seconds.\n' ...
      , T_dur(4) * 3600 + T_dur(5) * 60 + T_dur(6));

% code

[David Young]

The code is hard to read because lines are run together. You can format it using the "{} Code" button.

Do you have experience of some other language? That is, are you asking about the MATLAB-specific features of the code, or do you need features like if and for which are common to most languages explaining as well?

It's quite a big ask to explain all of a lengthy chunk of code like this. Can you perhaps focus on a part of it that you're finding particularly hard to understand?

[Azzi Abdelmalek]

Can you ask a more precise question?

[Echo]

is that clearer now?

[Echo]

Re: David,

I used MS C++ before,

And for this program, I knew the function of the program but I am not understanding the flow of ideas and the codes. So, I want to ask about this.

Answer 1

[Image Analyst]

Since your "program files are given by my supervisor " he would be the one to ask, rather than with us, don't you think?

[Echo]

I know what you mean, but actually, I ask many times and I still don't understand. Besides, his codes are not good also.

[John D'Errico]

Then you need to start learning MATLAB, AND you need to start asking more and better questions of your supervisor. Be persistent, as that is his job. If his codes are not good, then how do you expect us to know what he is doing from poorly written code? And if you don't know MATLAB, then how do you know his code is not good anyway? From what I've seen, I see many comments that explain nearly every line of code.

[Echo]

Yes, the comments are typed by him because I don't quite understand the flow and I asked him for this. So, he typed the comments for me.

And basically, I knew some of the MATLAB functions but i haven't type a program before.

[Image Analyst]

That was very nice of him to help you by doing that, though he should have put the comments in as the code was being written originally - that's what the best programmers do. I do that. Sounds like he's very willing to help you until you fully understand his code, and I'd take advantage of his generosity.

Answer 2

[David Young]

For someone coming from the C++ background, I suspect one unfamiliar aspect of the code is the colon operator (:), which generates a vector containing a sequence of numerical values. For example,

3:6

produces the vector

[3 4 5 6]

Give the command "doc colon" to get more information and examples.

Then you need to become familiar with the idea that operators like + and * can be applied to the whole of a vector or matrix, so that what would be a loop in C++ becomes a single line of code. For example

[3 4 5] + 10

produces the vector

[13 14 15]

Again, you'll need to follow this up in the documentation.

On the overall structure, it looks like you call the script Sim.m to run the program, and it in turn calls Sim_FIFO_MDc.m. Calling scripts in this way is a bit like function call, except without input and output parameters and with a shared workspace for variables. Otherwise, the control structures ("for" and "if") are much as in other languages.

One thing that may be misleading is the presence of all the "global" declarations. These are unnecessary, since scripts all share a single workspace, so all their variables are shared.

The comments in the code look very helpful. You should be able to make progress by going through the code a line at a time, looking at the documentation for any construct you don't understand. You can even execute it a line at a time using the debugger, and then you can see exactly what it does and what the variable values are at each stage.

[Echo]

Wow, it is very nice that you tell me these. I am getting familiar with this now. But can we have more conversation to ask you about MATLAB via other communication channel like skype or any other else?

[David Young]

Sorry. I'm happy to answer a few questions here, but I'm afraid I don't have time to do individual help.