How to generate 100 sample points of a 3-state Markov chain with initial prob vector and transition probability matrix using inverse transform method on matlab.

P   = [0.3 0.1 0.1; 0.3 0.9 0.1; 0.4 0 0.8]; % Transition Matrix
P0  = [0.1 0.9 0]; %initial probability Vector

Then I dont know how to do the rest. Please help me.

Answer 1

[Rick Rosson]

Here's a start:

   P = [0.3 0.1 0.1; 0.3 0.9 0.1; 0.4 0 0.8]; % Transition Matrix 
   P0 = [0.1 0.9 0]; %initial probability Vector

   C0 = cumsum(P0);
   M = length(P0);

   N = 100;
   x = nan(N,1);
   x(1) = (M + 1) - sum(rand <= C0);

   for k = 2:N
       x(k) = ...
   end

[bmp_sq]

Thank you!

[bmp_sq]

what does nan(N,1) function do btw? Can you explain. I never use matlab before. thanks.

[Stephen23]

@jndmwx: then you need to learn to use the documentation. You can search the documentation using any good internet search engine. I just searched for "MATLAB nan", and this is the first result that was returned:

http://www.mathworks.com/help/matlab/ref/nan.html

"For example, NaN(3,4) returns a 3-by-4 array of NaN values".