accidental bug fix

cast constants to tensors

Author: Junpeng Lao

pymc3/distributions/continuous.py

@@ -376,7 +376,7 @@ def __init__(self, mu=None, lam=None, phi=None, alpha=0., *args, **kwargs):
         self.alpha = alpha = tt.as_tensor_variable(alpha)
         self.mu = mu = tt.as_tensor_variable(mu)
         self.lam = lam = tt.as_tensor_variable(lam)
-        self.phi = phi =tt.as_tensor_variable(phi)
+        self.phi = phi = tt.as_tensor_variable(phi)
 
         self.mean = self.mu + self.alpha
         self.mode = self.mu * (tt.sqrt(1. + (1.5 * self.mu / self.lam)**2)
@@ -1501,11 +1501,9 @@ def __init__(self, lower=0, upper=1, c=0.5,
                  *args, **kwargs):
         super(Triangular, self).__init__(*args, **kwargs)
 
-        self.c = c
-        self.lower = lower
-        self.upper = upper
-        self.mean = c
-        self.median = self.mean
+        self.median = self.mean = self.c = c  = tt.as_tensor_variable(c)
+        self.lower = lower = tt.as_tensor_variable(lower)
+        self.upper = upper = tt.as_tensor_variable(upper)
 
     def random(self, point=None, size=None):
         c, lower, upper = draw_values([self.c, self.lower, self.upper],

pymc3/distributions/discrete.py

@@ -206,8 +206,8 @@ class DiscreteWeibull(Discrete):
     def __init__(self, q, beta, *args, **kwargs):
         super(DiscreteWeibull, self).__init__(*args, defaults=['median'], **kwargs)
 
-        self.q = q
-        self.beta = beta
+        self.q = q = tt.as_tensor_variable(q)
+        self.beta = beta = tt.as_tensor_variable(beta)
 
         self.median = self._ppf(0.5)
 

pymc3/distributions/mixture.py

@@ -42,7 +42,7 @@ class Mixture(Distribution):
     def __init__(self, w, comp_dists, *args, **kwargs):
         shape = kwargs.pop('shape', ())
 
-        self.w = w
+        self.w = w = tt.as_tensor_variable(w)
         self.comp_dists = comp_dists
 
         defaults = kwargs.pop('defaults', [])

pymc3/distributions/multivariate.py

@@ -594,9 +594,9 @@ def __init__(self, nu, V, *args, **kwargs):
                       'on the issues surrounding the Wishart see here: '
                       'https://github.com/pymc-devs/pymc3/issues/538.',
                       UserWarning)
-        self.nu = nu
-        self.p = p = V.shape[0]
-        self.V = V
+        self.nu = nu = tt.as_tensor_variable(nu)
+        self.p = p = tt.as_tensor_variable(V.shape[0])
+        self.V = V = tt.as_tensor_variable(V)
         self.mean = nu * V
         self.mode = tt.switch(1 * (nu >= p + 1),
                               (nu - p - 1) * V,
@@ -840,8 +840,8 @@ class LKJCholeskyCov(Continuous):
        http://math.stackexchange.com/q/130026
     """
     def __init__(self, eta, n, sd_dist, *args, **kwargs):
-        self.n = n
-        self.eta = eta
+        self.n = n = tt.as_tensor_variable(n)
+        self.eta = eta = tt.as_tensor_variable(eta)
 
         if 'transform' in kwargs:
             raise ValueError('Invalid parameter: transform.')
@@ -943,13 +943,13 @@ def __init__(self, eta=None, n=None, p=None, transform='interval', *args, **kwar
                           'dimension parameter p -> n. Please update your code. '
                           'Automatically re-assigning parameters for backwards compatibility.',
                           DeprecationWarning)
-            self.n = p
-            self.eta = n
+            self.n = p = tt.as_tensor_variable(p)
+            self.eta = n = tt.as_tensor_variable(n)
             eta = self.eta
             n = self.n
         elif (n is not None) and (eta is not None) and (p is None):
-            self.n = n
-            self.eta = eta
+            self.n = n = tt.as_tensor_variable(n)
+            self.eta = eta = tt.as_tensor_variable(eta)
         else:
             raise ValueError('Invalid parameter: please use eta as the shape parameter and '
                              'n as the dimension parameter.')

pymc3/distributions/timeseries.py

@@ -30,10 +30,10 @@ class AR1(distribution.Continuous):
 
     def __init__(self, k, tau_e, *args, **kwargs):
         super(AR1, self).__init__(*args, **kwargs)
-        self.k = k
-        self.tau_e = tau_e
+        self.k = k = tt.as_tensor_variable(k)
+        self.tau_e = tau_e = tt.as_tensor_variable(tau_e)
         self.tau = tau_e * (1 - k ** 2)
-        self.mode = 0.
+        self.mode = tt.as_tensor_variable(0.)
 
     def logp(self, x):
         k = self.k
@@ -75,11 +75,11 @@ class GaussianRandomWalk(distribution.Continuous):
     def __init__(self, tau=None, init=continuous.Flat.dist(), sd=None, mu=0.,
                  *args, **kwargs):
         super(GaussianRandomWalk, self).__init__(*args, **kwargs)
-        self.tau = tau
-        self.sd = sd
-        self.mu = mu
+        self.tau = tau = tt.as_tensor_variable(tau)
+        self.sd = sd = tt.as_tensor_variable(sd)
+        self.mu = mu = tt.as_tensor_variable(mu)
         self.init = init
-        self.mean = 0.
+        self.mean = tt.as_tensor_variable(0.)
 
     def logp(self, x):
         tau = self.tau
@@ -129,11 +129,11 @@ def __init__(self, omega=None, alpha_1=None, beta_1=None,
                  initial_vol=None, *args, **kwargs):
         super(GARCH11, self).__init__(*args, **kwargs)
 
-        self.omega = omega
-        self.alpha_1 = alpha_1
-        self.beta_1 = beta_1
+        self.omega = omega = tt.as_tensor_variable(omega)
+        self.alpha_1 = alpha_1 = tt.as_tensor_variable(alpha_1)
+        self.beta_1 = beta_1 = tt.as_tensor_variable(beta_1)
         self.initial_vol = initial_vol
-        self.mean = 0
+        self.mean = tt.as_tensor_variable(0.)
 
     def get_volatility(self, x):
         x = x[:-1]
@@ -179,7 +179,7 @@ class EulerMaruyama(distribution.Continuous):
     """
     def __init__(self, dt, sde_fn, sde_pars, *args, **kwds):
         super(EulerMaruyama, self).__init__(*args, **kwds)
-        self.dt = dt
+        self.dt = dt = tt.as_tensor_variable(dt)
         self.sde_fn = sde_fn
         self.sde_pars = sde_pars
 
@@ -221,9 +221,9 @@ def __init__(self, mu=0., cov=None, init=continuous.Flat.dist(),
         if cov.ndim != 2:
             raise ValueError('cov must be two dimensional.')
         self.cov = cov
-        self.mu = mu
+        self.mu = mu = tt.as_tensor_variable(mu)
         self.init = init
-        self.mean = 0.
+        self.mean = tt.as_tensor_variable(0.)
 
     def logp(self, x):
         cov = self.cov
@@ -263,10 +263,10 @@ class MvStudentTRandomWalk(distribution.Continuous):
     def __init__(self, nu, mu=0., cov=None, init=continuous.Flat.dist(),
                  *args, **kwargs):
         super(MvStudentTRandomWalk, self).__init__(*args, **kwargs)
-        self.mu = mu
-        self.nu = nu
+        self.mu = mu = tt.as_tensor_variable(mu)
+        self.nu = nu = tt.as_tensor_variable(nu)
         self.init = init
-        self.mean = 0.
+        self.mean = tt.as_tensor_variable(0.)
 
         if cov is None:
             raise ValueError('A covariance matrix must be provided as cov argument.')