Text

#Matriz a llenar con las probabilidades de transición
m = numpy.zeros((452, 452), numpy.float32)

with open('/Users/agutierrez/Documents/maestria/statistical-computing/compstat2016/proyecto/2016-09.csv', 'rb') as csvfile:
    rows = csv.reader(csvfile, delimiter=';')
    rows.next()
    for row in rows:
        # I don't want to consider odd station numbers such as 1002
    	if int(row[3]) > 500 or int(row[6]) > 500:
            pass
        # If we hit this else
        else:
        	station_in = int(row[3]) - 1
        	station_out = int(row[6]) - 1

        	m[station_in][station_out] = m[station_in][station_out] + 1.0

#Convertimos a probabilidades
m_normalized = m / m.sum(axis=1,keepdims=True)

#Elevando la matríz de transición para obtener la estacionaria
m_powered = numpy.linalg.matrix_power(m_normalized, 100)

38%

62%

Edad

Modo de transporte

13%

               de los viajes se realizan sólo en EcoBici

               de los viajes se
combinan con otros modos de transporte

 

87%

\left( \begin{array}{ccc} 2 & 0 & 8 \\ 6 & 1 & 3 \\ 5 & 7 & 0 \end{array} \right)
(208613570)\left( \begin{array}{ccc} 2 & 0 & 8 \\ 6 & 1 & 3 \\ 5 & 7 & 0 \end{array} \right)
\left( \begin{array}{ccc} 0.2 & 0 & 0.8 \\ 0.6 & 0.3 & 0.1 \\ 0.42 & 0.58 & 0 \end{array} \right)
(0.200.80.60.30.10.420.580)\left( \begin{array}{ccc} 0.2 & 0 & 0.8 \\ 0.6 & 0.3 & 0.1 \\ 0.42 & 0.58 & 0 \end{array} \right)
\left( \begin{array}{cc} 1-p & p \\ q & 1-q \end{array} \right)
(1ppq1q)\left( \begin{array}{cc} 1-p & p \\ q & 1-q \end{array} \right)
\lim_{n\rightarrow\infty}\left( \begin{array}{cc} p & 1-p \\ 1-q & q \end{array} \right)^n = \left( \begin{array}{cc} \frac{q}{p+q} & \frac{p}{p+q}\\ \frac{q}{p+q} & \frac{p}{p+q}\end{array} \right)
limn(p1p1qq)n=(qp+qpp+qqp+qpp+q)\lim_{n\rightarrow\infty}\left( \begin{array}{cc} p & 1-p \\ 1-q & q \end{array} \right)^n = \left( \begin{array}{cc} \frac{q}{p+q} & \frac{p}{p+q}\\ \frac{q}{p+q} & \frac{p}{p+q}\end{array} \right)
P(x)\approx \frac{n_x}{n_i}
P(x)nxniP(x)\approx \frac{n_x}{n_i}

Cadenas de Markov

Una cadena de Markov es una serie de eventos, en la cual la probabilidad de que ocurra un evento depende del evento inmediato anterior.

\lim_{n\rightarrow\infty}\left( \begin{array}{ccc} 0.2 & 0 & 0.8 \\ 0.6 & 0.3 & 0.1 \\ 0.42 & 0.58 & 0 \end{array} \right)^n=\left( \begin{array}{ccc} 0.38 & 0.28 & 0.34 \\ 0.38 & 0.28 & 0.34 \\ 0.38 & 0.28 & 0.34 \end{array} \right)
limn(0.200.80.60.30.10.420.580)n=(0.380.280.340.380.280.340.380.280.34)\lim_{n\rightarrow\infty}\left( \begin{array}{ccc} 0.2 & 0 & 0.8 \\ 0.6 & 0.3 & 0.1 \\ 0.42 & 0.58 & 0 \end{array} \right)^n=\left( \begin{array}{ccc} 0.38 & 0.28 & 0.34 \\ 0.38 & 0.28 & 0.34 \\ 0.38 & 0.28 & 0.34 \end{array} \right)

Cadenas de Markov

By Amaury Gutiérrez

Cadenas de Markov

  • 576