https://github.com/GPflow/GPflow
Revision 5c78295454d92838240b376e8c0f2ac6891dbdfd authored by alexggmatthews on 31 May 2016, 14:57:53 UTC, committed by alexggmatthews on 31 May 2016, 14:57:53 UTC
1 parent af1212a
Raw File
Tip revision: 5c78295454d92838240b376e8c0f2ac6891dbdfd authored by alexggmatthews on 31 May 2016, 14:57:53 UTC
Reverting logistic transformation.
Tip revision: 5c78295
priors.py
from . import densities
import tensorflow as tf
import numpy as np
from .param import Parameterized

class Prior(Parameterized):
    def logp(self, x):
        """
        The log density of the prior as x

        All priors (for the moment) are univariate, so if x is a vector or an array, this is the sum of the log densities.
        """
        raise NotImplementedError

    def __str__(self):
        """
        A short string to describe the prior at print time
        """
        raise NotImplementedError

class Gaussian(Prior):
    def __init__(self, mu, var):
        Prior.__init__(self)
        self.mu, self.var = np.atleast_1d(np.array(mu, np.float64)), np.atleast_1d(np.array(var, np.float64))
    def logp(self, x):
        return tf.reduce_sum(densities.gaussian(x, self.mu, self.var))
    def __str__(self):
        return "N("+str(self.mu) + "," + str(self.var) + ")"

class Gamma(Prior):
    def __init__(self, shape, scale):
        Prior.__init__(self)
        self.shape, self.scale = np.atleast_1d(np.array(shape, np.float64)), np.atleast_1d(np.array(scale, np.float64))
    def logp(self, x):
        return tf.reduce_sum(densities.gamma(self.shape, self.scale, x))
    def __str__(self):
        return "Ga("+str(self.shape) + "," + str(self.scale) + ")"
back to top