next up previous contents
Next: Continuous time Up: Examples Previous: Examples   Contents

Discrete time

Consider the classical Markov chain with two states :


\includegraphics[height=3cm,width=5cm]{markov1.eps}



$\displaystyle \mathbf{P}=
\left(\begin{tabular}{cc}
\par 1-a & a \\
b & 1-b \\
\end{tabular} \right)
$



where $ 0<a,b<1$, and $ a+b>0$.


This stochastic matrix admit as stationary distribution :

$\displaystyle \pi=(\frac b {a+b},\frac a {a+b})$




Take for example : $ a=\frac 1 2$ and $ b=\frac 1 3$, we obtain the transition matrix :

$\displaystyle \mathbf{P}=
\left(\begin{tabular}{cc}
$\frac 1 2$ & $\frac 1 2$ \\
$\frac 1 3$ & $\frac 2 3$ \\
\end{tabular} \right)
$

Look three importation possibilities in PSI :

-
'2states.hbf_r'.
-
'2states.hbf_c'.
-
'2states.marca'.



\fbox{\usebox{\fmbox}}

\fbox{\usebox{\fmbox}}

\fbox{\usebox{\fmbox}}




Note that it can be important to specify several decimal values to avoid numerical problems in Aliasing construction.

After precomputation, we obtain a '2states.simu' file (modulo row permutations) :



\fbox{\usebox{\fmbox}}



Florent Morata 2002-12-11