[priors] pickling priors (not working for Discriminative prior)

GPy/core/parameterization/priors.py

@@ -12,6 +12,11 @@ import weakref
 
 class Prior(object):
     domain = None
+    _instance = None
+    def __new__(cls, *args, **kwargs):
+        if not cls._instance or cls._instance.__class__ is not cls:
+            cls._instance = super(Prior, cls).__new__(cls, *args, **kwargs)
+        return cls._instance
 
     def pdf(self, x):
         return np.exp(self.lnpdf(x))
@@ -41,7 +46,7 @@ class Gaussian(Prior):
     domain = _REAL
     _instances = []
 
-    def __new__(cls, mu, sigma):  # Singleton:
+    def __new__(cls, mu=0, sigma=1):  # Singleton:
         if cls._instances:
             cls._instances[:] = [instance for instance in cls._instances if instance()]
             for instance in cls._instances:
@@ -69,20 +74,20 @@ class Gaussian(Prior):
     def rvs(self, n):
         return np.random.randn(n) * self.sigma + self.mu
 
-    def __getstate__(self):
-        return self.mu, self.sigma
-
-    def __setstate__(self, state):
-        self.mu = state[0]
-        self.sigma = state[1]
-        self.sigma2 = np.square(self.sigma)
-        self.constant = -0.5 * np.log(2 * np.pi * self.sigma2)
+#     def __getstate__(self):
+#         return self.mu, self.sigma
+#
+#     def __setstate__(self, state):
+#         self.mu = state[0]
+#         self.sigma = state[1]
+#         self.sigma2 = np.square(self.sigma)
+#         self.constant = -0.5 * np.log(2 * np.pi * self.sigma2)
 
 class Uniform(Prior):
     domain = _REAL
     _instances = []
 
-    def __new__(cls, lower, upper):  # Singleton:
+    def __new__(cls, lower=0, upper=1):  # Singleton:
         if cls._instances:
             cls._instances[:] = [instance for instance in cls._instances if instance()]
             for instance in cls._instances:
@@ -109,12 +114,12 @@ class Uniform(Prior):
     def rvs(self, n):
         return np.random.uniform(self.lower, self.upper, size=n)
 
-    def __getstate__(self):
-        return self.lower, self.upper
-
-    def __setstate__(self, state):
-        self.lower = state[0]
-        self.upper = state[1]
+#     def __getstate__(self):
+#         return self.lower, self.upper
+#
+#     def __setstate__(self, state):
+#         self.lower = state[0]
+#         self.upper = state[1]
 
 class LogGaussian(Gaussian):
     """
@@ -129,7 +134,7 @@ class LogGaussian(Gaussian):
     domain = _POSITIVE
     _instances = []
 
-    def __new__(cls, mu, sigma):  # Singleton:
+    def __new__(cls, mu=0, sigma=1):  # Singleton:
         if cls._instances:
             cls._instances[:] = [instance for instance in cls._instances if instance()]
             for instance in cls._instances:
@@ -158,7 +163,7 @@ class LogGaussian(Gaussian):
         return np.exp(np.random.randn(n) * self.sigma + self.mu)
 
 
-class MultivariateGaussian:
+class MultivariateGaussian(Prior):
     """
     Implementation of the multivariate Gaussian probability function, coupled with random variables.
 
@@ -171,7 +176,7 @@ class MultivariateGaussian:
     domain = _REAL
     _instances = []
 
-    def __new__(cls, mu, var):  # Singleton:
+    def __new__(cls, mu=0, var=1):  # Singleton:
         if cls._instances:
             cls._instances[:] = [instance for instance in cls._instances if instance()]
             for instance in cls._instances:
@@ -247,7 +252,7 @@ class Gamma(Prior):
     domain = _POSITIVE
     _instances = []
 
-    def __new__(cls, a, b):  # Singleton:
+    def __new__(cls, a=1, b=.5):  # Singleton:
         if cls._instances:
             cls._instances[:] = [instance for instance in cls._instances if instance()]
             for instance in cls._instances:
@@ -318,7 +323,7 @@ class InverseGamma(Gamma):
     """
     domain = _POSITIVE
     _instances = []
-    def __new__(cls, a, b): # Singleton:
+    def __new__(cls, a=1, b=.5): # Singleton:
         if cls._instances:
             cls._instances[:] = [instance for instance in cls._instances if instance()]
             for instance in cls._instances: