Merge 60c2529 into 57fa938

pgmpy/factors/__init__.py

@@ -6,6 +6,7 @@
 from .continuous.ContinuousFactor import ContinuousFactor
 from .continuous.JointGaussianDistribution import JointGaussianDistribution
 from .continuous.CanonicalFactor import CanonicalFactor
+from .continuous.LinearGaussianCPD import LinearGaussianCPD
 
 __all__ = ['Factor',
            'State',
@@ -19,4 +20,5 @@
            'ContinuousNode',
            'ContinuousFactor',
            'JointGaussianDistribution',
-           'CanonicalFactor']
+           'CanonicalFactor',
+           'LinearGaussianCPD']

pgmpy/factors/continuous/JointGaussianDistribution.py

@@ -99,7 +99,7 @@ def marginalize(self, variables, inplace=True):
         ----------
 
         variables: iterator
-                List of variables over which to marginalize.
+                List of variables over which marginalization is to be done.
 
         inplace: boolean
                 If inplace=True it will modify the distribution itself,

pgmpy/factors/continuous/LinearGaussianCPD.py

@@ -0,0 +1,133 @@
+# -*- coding: utf-8 -*-
+
+from __future__ import division
+
+import numpy as np
+from scipy.stats import multivariate_normal
+
+from pgmpy.extern import six
+from pgmpy.factors import ContinuousFactor
+
+
+class LinearGaussianCPD(ContinuousFactor):
+    u"""
+    For, X -> Y the Linear Gaussian model assumes that the mean
+    of Y is a linear function of mean of X and the variance of Y does
+    not depend on X.
+
+    For example,
+    p(Y|X) = N(-2x + 0.9 ; 1)
+    Here, x is the mean of the variable X.
+
+    Let Y be a continuous variable with continuous parents
+    X1 ............ Xk . We say that Y has a linear Gaussian CPD
+    if there are parameters β0,.........βk and σ2 such that,
+
+    p(Y |x1.......xk) = N(β0 + x1*β1 + ......... + xk*βk ; σ2)
+
+    In vector notation,
+
+    p(Y |x) = N(β0 + β.T * x ; σ2)
+
+    """
+    def __init__(self, variable, beta_0, variance, evidence=[], beta_vector=[]):
+        """
+        Parameters
+        ----------
+
+        variable: any hashable python object
+            The variable whose CPD is defined.
+
+        beta_0: int, float
+            Represents the constant term in the linear equation.
+
+        variance: int, float
+            The variance of the variable defined.
+
+        evidence: iterable of any hashabale python objects
+            An iterable of the parents of the variable. None
+            if there are no parents.
+
+        beta_vector: iterable of int or float
+            An iterable representing the coefficient vector of the linear equation.
+
+        Examples
+        --------
+
+        # For P(Y| X1, X2, X3) = N(-2x1 + 3x2 + 7x3 + 0.2; 9.6)
+
+        >>> cpd = LinearGaussianCPD('Y', 0.2, 9.6, ['X1', 'X2', 'X3'], [-2, 3, 7])
+        >>> cpd.variable
+        'Y'
+        >>> cpd.variance
+        9.6
+        >>> cpd.evidence
+        ['x1', 'x2', 'x3']
+        >>> cpd.beta_vector
+        [-2, 3, 7]
+        >>> cpd.beta_0
+        0.2
+
+        """
+        self.variable = variable
+        self.beta_0 = beta_0
+        self.variance = variance
+
+        if len(evidence) != len(beta_vector):
+            raise ValueError("The number of variables in evidence must be equal to the "
+                             "length of the beta vector.")
+
+        self.evidence = evidence
+        self.beta_vector = np.asarray(beta_vector)
+
+        variables = [variable] + evidence
+        super(LinearGaussianCPD, self).__init__(variables, None)
+
+    @property
+    def pdf(self):
+
+        def _pdf(*args):
+            # The first element of args is the value of the variable on which CPD is defined
+            # and the rest of the elements give the mean values of the parent variables.
+            mean = sum([arg * coeff for (arg, coeff) in zip(args[1:], self.beta_vector)]) + self.beta_0
+            return multivariate_normal.pdf(args[0], np.array(mean), np.array([[self.variance]]))
+
+        return _pdf
+
+    def copy(self):
+        """
+        Return a copy of the distribution.
+
+        Returns
+        -------
+        LinearGaussianCPD: copy of the distribution
+
+        Examples
+        --------
+        >>> import numpy as np
+        >>> from pgmpy.factors import LinearGaussianCPD
+        >>> cpd = LinearGaussianCPD('Y', 0.2, 9.6, ['X1', 'X2', 'X3'], [-2, 3, 7])
+        >>> copy_cpd = cpd.copy()
+        >>> copy_cpd.variable
+        'Y'
+        >>> copy_cpd.evidence
+        ['X1', 'X2', 'X3']
+ 
+        """
+        copy_cpd = LinearGaussianCPD(self.variable, self.beta_0, self.variance,
+                                     list(self.evidence), self.beta_vector)
+
+        return copy_cpd
+
+    def __str__(self):
+        if self.evidence and list(self.beta_vector):
+            # P(Y| X1, X2, X3) = N(-2*X1_mu + 3*X2_mu + 7*X3_mu; 0.2)
+            rep_str = "P(" + str(self.variable) + "| " + ", ".join([str(var) for var in self.evidence]) + ") = " +\
+                      "N(" + " + ".join(["{coeff}*{parent}_mu".format(coeff=coeff, parent=parent)
+                                        for coeff, parent in zip(self.beta_vector,
+                                                                 self.evidence)]) + "; " + str(self.beta_0) + ")"
+        else:
+            # P(X) = N(1, 4)
+            rep_str = "P({X}) = N({beta_0}; {variance})".format(X=str(self.variable), beta_0=str(self.beta_0),
+                                                                variance=str(self.variance))
+        return rep_str

pgmpy/models/BayesianModel.py

@@ -177,6 +177,10 @@ def get_cpds(self, node=None):
             The node whose CPD we want. If node not specified returns all the
             CPDs added to the model.
 
+        Returns
+        -------
+        A list of TabularCPDs.
+
         Examples
         --------
         >>> from pgmpy.models import BayesianModel
@@ -204,7 +208,7 @@ def remove_cpds(self, *cpds):
 
         Parameters
         ----------
-        *cpds: TabularCPD, TreeCPD, RuleCPD object
+        *cpds: TabularCPD object
             A CPD object on any subset of the variables of the model which
             is to be associated with the model.
 

pgmpy/models/__init__.py

@@ -7,6 +7,7 @@
 from .DynamicBayesianNetwork import DynamicBayesianNetwork
 from .MarkovChain import MarkovChain
 from .NaiveBayes import NaiveBayes
+from .continuous.LinearGaussianBayesianNetwork import LinearGaussianBayesianNetwork
 
 __all__ = ['BayesianModel',
            'NoisyOrModel',
@@ -16,4 +17,5 @@
            'ClusterGraph',
            'DynamicBayesianNetwork',
            'MarkovChain',
-           'NaiveBayes']
+           'NaiveBayes',
+           'LinearGaussianBayesianNetwork']