def discrete_shannon_entropy(x, n_bins, _range=None):
from R import digamma #@UnresolvedImport
hist = density(x, n_bins, _range, steps=0, hist=1)
x = hist[:, 1]
v = Numeric.sum(x)
s = shape(x)
z = zeros(s, Float)
for i in range(s[0]):
z[i] = digamma(x[i] + 1)
return -Numeric.sum(x / v * (z - digamma(v + 1)))
def discrete_shannon_entropy(x, n_bins, _range = None):
from R import digamma #@UnresolvedImport
hist = density(x, n_bins, _range, steps = 0, hist = 1)
x = hist[:,1]
v = Numeric.sum(x)
s = shape(x)
z = zeros(s, Float)
for i in range(s[0]):
z[i] = digamma(x[i] + 1)
return - Numeric.sum(x / v * (z - digamma(v + 1)))
def mutualInformation(counts):
"""
bayesian estimator for the mutual information
between two series of discrete numbers.
'counts': matrix of counts
"""
from R import digamma #@UnresolvedImport
counts = Numeric.array(counts).astype(Float32)
v = Numeric.sum(Numeric.sum(counts))
r = digamma(v + 1)
s = shape(counts)
z = Numeric.zeros(shape(counts), Float)
for i in range(s[0]):
for j in range(s[1]):
z[i, j] = digamma(counts[i, j] + 1)
a = Numeric.sum(Numeric.sum(counts / v * (z - r)))
m1 = Numeric.sum(counts, 1)
m0 = Numeric.sum(counts)
z0 = Numeric.zeros(shape(m0), Float)
z1 = Numeric.zeros(shape(m1), Float)
for i in range(len(m0)):
z0[i] = digamma(m0[i] + 1)
for i in range(len(z1)):
z1[i] = digamma(m1[i] + 1)
b = Numeric.sum(m0 / v * (z0 - r))
c = Numeric.sum(m1 / v * (z1 - r))
return a - b - c
def mutualInformation(counts):
"""
bayesian estimator for the mutual information
between two series of discrete numbers.
'counts': matrix of counts
"""
from R import digamma #@UnresolvedImport
counts = Numeric.array(counts).astype(Float32)
v = Numeric.sum(Numeric.sum(counts))
r = digamma(v + 1)
s = shape(counts)
z = Numeric.zeros(shape(counts), Float)
for i in range(s[0]):
for j in range(s[1]):
z[i,j] = digamma(counts[i,j] + 1)
a = Numeric.sum(Numeric.sum(counts / v * (z - r)))
m1 = Numeric.sum(counts, 1)
m0 = Numeric.sum(counts)
z0 = Numeric.zeros(shape(m0), Float)
z1 = Numeric.zeros(shape(m1), Float)
for i in range(len(m0)):
z0[i] = digamma(m0[i] + 1)
for i in range(len(z1)):
z1[i] = digamma(m1[i] + 1)
b = Numeric.sum(m0 / v * (z0 - r))
c = Numeric.sum(m1 / v * (z1 - r))
return a - b - c
