def Likelihood(self, data, hypo):
"""Computes the likelihood of the data under the hypothesis.
hypo:
data:
"""
std = 30
meanx, meany = hypo
x, y = data
like = thinkbayes2.EvalNormalPdf(x, meanx, std)
like *= thinkbayes2.EvalNormalPdf(y, meany, std)
return like
def RenderPdf(mu, sigma, n=101):
"""Makes xs and ys for a normal PDF with (mu, sigma).
n: number of places to evaluate the PDF
"""
xs = numpy.linspace(mu - 4 * sigma, mu + 4 * sigma, n)
ys = [thinkbayes2.EvalNormalPdf(x, mu, sigma) for x in xs]
return xs, ys
def Likelihood(self, data, hypo):
"""Computes the likelihood of the data under the hypothesis.
hypo: the percentage of voters who you hypothesize to favor your candidate
data: tuple (mean, standard deviation, measurement)
"""
mean, std_dev, measurement = data
error = measurement - hypo
likelihood = thinkbayes2.EvalNormalPdf(error, mean, std_dev)
return likelihood
def Likelihood(self, data, hypo):
'''Computes the likelihood of the data under the hypothesis
data: tuple of (number of sensing, number of intuitive)
hypo: mu and sigma pair describing the distribution of learning styles'''
mu, sigma = hypo
x = data
like = thinkbayes2.EvalNormalPdf(x, mu, sigma)
return like
def Likelihood(self, data, hypo):
"""Computes the likelihood of the data under the hypothesis.
hypo: fraction of the population that supports your candidate
data: poll results
"""
bias, std, result = data
error = result - hypo
like = thinkbayes2.EvalNormalPdf(error, bias, std)
return like
def Likelihood(self, data, hypo):
"""Computes the likelihood of the data under the hypothesis.
hypo: x, y
data: x, y
"""
measured_x, measured_y = data
actual_x, actual_y = hypo
error_x = measured_x - actual_x
error_y = measured_y - actual_y
prob_x = thinkbayes2.EvalNormalPdf(error_x, 0, 30)
prob_y = thinkbayes2.EvalNormalPdf(error_y, 0, 30)
like = prob_x * prob_y
return like
def Likelihood(self, data, hypo):
'''
Likelihood of the data under the hypothesis.
hypo: fraction of the population
data: poll results
'''
bias, std, result = data
error = result - hypo
like = thinkbayes2.EvalNormalPdf(error, bias, std)
return like
def Likelihood(self, data, hypo):
"""Computes the likelihood of the data under the hypothesis.
hypo:
data:
"""
a_hypo = hypo
mean, std, measurement = data
e_hypo = measurement - a_hypo
like = thinkbayes2.EvalNormalPdf(e_hypo, mean, std)
return like
def Likelihood(self, data, hypo):
"""
Calculate the likelihood of a hypo given the data
data: [(Timestamp, 12.142122), ...] pandas dataframe of records in the form of (date, record in mph) tuples
hypo: (alpha, beta, sigma)
"""
alpha, beta, sigma = hypo
total_likelihood = 1
for i, row in enumerate(data.values):
date = row[1].value # ms
measured_mph = row[0]
predicted_mph = alpha + beta * date
error = measured_mph - predicted_mph
total_likelihood *= thinkbayes2.EvalNormalPdf(error, mu=0, sigma=sigma)
return total_likelihood