add sympy support; which can transfer your network into sympy expression;

add visualize in genome;
add related tests.

tensorneat/algorithm/neat/genome/recurrent.py

@@ -84,3 +84,6 @@ class RecurrentGenome(BaseGenome):
             return vals[self.output_idx]
         else:
             return self.output_transform(vals[self.output_idx])
+
+    def sympy_func(self, state, network, precision=3):
+        raise ValueError("Sympy function is not supported for Recurrent Network!")