Allow loading more nfunc path (like class method) and bug fixes.

theano/tensor/elemwise.py

@@ -392,7 +392,7 @@ second dimension
             inplace_pattern = frozendict({})
         self.name = name
         self.scalar_op = scalar_op
-        self.inplace_pattern = frozendict(inplace_pattern)
+        self.inplace_pattern = inplace_pattern
         self.destroy_map = dict((o, [i]) for o, i in self.inplace_pattern.items())
 
         self.ufunc = None
@@ -400,16 +400,7 @@ second dimension
         if nfunc_spec is None:
             nfunc_spec = getattr(scalar_op, 'nfunc_spec', None)
         self.nfunc_spec = nfunc_spec
-        if nfunc_spec:
-            self.nfunc = getattr(np, nfunc_spec[0], None)
-            if self.nfunc is None:
-                # Not inside NumPy. So probably another package like scipy.
-                symb = 'scipy.special.erfinv'.split(".")
-                module = __import__('.'.join(symb[:-1]))
-                for sub in symb[1:-1]:
-                    module = getattr(module, sub)
-                self.nfunc = getattr(module, symb[-1])
-
+        self.__setstate__(self.__dict__)
         super(Elemwise, self).__init__(openmp=openmp)
 
     def __getstate__(self):
@@ -425,7 +416,19 @@ second dimension
         self.nfunc = None
         self.inplace_pattern = frozendict(self.inplace_pattern)
         if getattr(self, 'nfunc_spec', None):
-            self.nfunc = getattr(np, self.nfunc_spec[0])
+            self.nfunc = getattr(np, self.nfunc_spec[0], None)
+            if self.nfunc is None:
+                # Not inside NumPy. So probably another package like scipy.
+                symb = self.nfunc_spec[0].split(".")
+                for idx in range(1, len(self.nfunc_spec[0])):
+                    try:
+                        module = __import__('.'.join(symb[:idx]))
+                    except ImportError:
+                        break
+                for sub in symb[1:]:
+                    module = getattr(module, sub)
+                self.nfunc = module
+
         elif 0 < self.scalar_op.nin < 32:
             self.ufunc = np.frompyfunc(self.scalar_op.impl,
                                        self.scalar_op.nin,