made chi2sf binary, fixed reviewer comments

theano/scalar/basic_scipy.py

@@ -2,7 +2,7 @@
 #as scipy is not always available, we treat them separatly
 import numpy
 
-from theano.scalar.basic import (UnaryScalarOp,
+from theano.scalar.basic import (UnaryScalarOp, BinaryScalarOp,
                                  exp, upgrade_to_float,
                                  float_types)
 from theano.scalar.basic import (upgrade_to_float_no_complex,
@@ -14,8 +14,6 @@ try:
     import scipy.special
     import scipy.stats
     imported_scipy_special = True
-# note: the Erf functions check for this but others do not.  Should they?
-
 # Importing scipy.special may raise ValueError.
 # See http://projects.scipy.org/scipy/ticket/1739
 except (ImportError, ValueError):
@@ -155,7 +153,10 @@ class Gamma(UnaryScalarOp):
         return scipy.special.gamma(x)
 
     def impl(self, x):
+        if imported_scipy_special:
             return Gamma.st_impl(x)
+        else:
+            super(Gamma, self).impl(x)
 
     def grad(self, (x, ), (gz, )):
         return gz * gamma(x) * psi(x),
@@ -182,8 +183,10 @@ class GammaLn(UnaryScalarOp):
         return scipy.special.gammaln(x)
 
     def impl(self, x):
+        if imported_scipy_special:
             return GammaLn.st_impl(x)
-
+        else:
+            super(GammaLn, self).impl(x)
     def grad(self, inp, grads):
         x, = inp
         gz, = grads
@@ -214,7 +217,10 @@ class Psi(UnaryScalarOp):
         return scipy.special.psi(x)
 
     def impl(self, x):
+        if imported_scipy_special:
             return Psi.st_impl(x)
+        else:
+            super(Psi, self).impl(x)
 
     def grad(self, inputs, outputs_gradients):
         raise NotImplementedError()
@@ -283,8 +289,7 @@ DEVICE double _psi(double x){
         return hash(type(self))
 psi = Psi(upgrade_to_float, name='psi')
 
-from theano.scalar import ScalarOp #necessary since chi2sf takes two inputs
-class Chi2SF(ScalarOp):
+class Chi2SF(BinaryScalarOp):
     """
     Compute (1 - chi2_cdf(x))
         ie. chi2 pvalue (chi2 'survival function')
@@ -294,11 +299,23 @@ class Chi2SF(ScalarOp):
     def st_impl(x, k):
         return scipy.stats.chi2.sf(x, k)
     def impl(self, x, k):
+        if imported_scipy_special:
             return Chi2SF.st_impl(x, k)
-    
+        else:
+            super(Chi2SF, self).impl(x, k)
     def c_support_code(self):
         return(
                """
+                   //For GPU support
+                   #ifdef __CUDACC__
+                   #define DEVICE __device__
+                   #else
+                   #define DEVICE
+                   #endif
+                   
+                   #ifndef _CHI2FUNCDEFINED
+                   #define _CHI2FUNCDEFINED
+                   
                    /*----------------------------------------------------------------------
                    File    : gamma.c
                    Contents: computation of the (incomplete/regularized) gamma function
@@ -534,7 +551,7 @@ class Chi2SF(ScalarOp):
                    
                    //ebuchman: this function is equivalent to scipy.stats.chi2.sf
                    //it's the pvalue (survival function) of a chi2 distribution
-                   double Chi2SF (double k, double x)
+                   DEVICE double Chi2SF (double k, double x)
                    {
                    return 1 - GammaP(k/2., x/2.);
                    }
@@ -546,8 +563,9 @@ class Chi2SF(ScalarOp):
         x, k = inp
         z, = out
         if node.inputs[0].type in float_types:
+            dtype = z.dtype
             return """%(z)s =
-                Chi2SF(%(k)s, %(x)s);""" % locals()
+                (%(dtype)s)Chi2SF(%(k)s, %(x)s);""" % locals()
         raise NotImplementedError('only floatingpoint is implemented')
     
     def __eq__(self, other):

theano/tensor/tests/test_basic.py

@@ -1555,7 +1555,7 @@ PsiInplaceTester = makeBroadcastTester(
 #chi2sf takes two inputs, a value (x) and a degrees of freedom (k).
 # not sure how to deal with that here...
     
-_good_broadcast_unary_psi = dict(
+_good_broadcast_unary_chi2sf = dict(
      normal=(rand_ranged(1, 10, (2, 3)),),
      empty=(numpy.asarray([]),),)