fix: torch export with constant arguments

pysr/export_torch.py

@@ -97,8 +97,8 @@ def __init__(self, *, expr, _memodict, _func_lookup, **kwargs):
                     self._torch_func = lambda: self._value
                     self._args = ()
                 elif issubclass(expr.func, sympy.Rational):
-                    # This is some fraction fixed in the operator.
-                    self._value = float(expr)
+                    # Includes Integer, since Integer is a subclass of Rational
+                    self.register_buffer("_value", torch.tensor(float(expr)))
                     self._torch_func = lambda: self._value
                     self._args = ()
                 elif issubclass(expr.func, sympy.UnevaluatedExpr):
@@ -111,15 +111,9 @@ def __init__(self, *, expr, _memodict, _func_lookup, **kwargs):
                     self.register_buffer("_value", torch.tensor(float(expr.args[0])))
                     self._torch_func = lambda: self._value
                     self._args = ()
-                elif issubclass(expr.func, sympy.Integer):
-                    # Can get here if expr is one of the Integer special cases,
-                    # e.g. NegativeOne
-                    self._value = int(expr)
-                    self._torch_func = lambda: self._value
-                    self._args = ()
                 elif issubclass(expr.func, sympy.NumberSymbol):
                     # Can get here from exp(1) or exact pi
-                    self._value = float(expr)
+                    self.register_buffer("_value", torch.tensor(float(expr)))
                     self._torch_func = lambda: self._value
                     self._args = ()
                 elif issubclass(expr.func, sympy.Symbol):

pysr/test/test_torch.py

@@ -192,6 +192,32 @@ def test_issue_656(self):
             decimal=3,
         )
 
+    def test_constant_arguments(self):
+        # Test that functions with constant arguments work correctly
+        # Regression test for https://github.com/MilesCranmer/PySR/issues/656
+        test_cases = [
+            (pysr.export_sympy.pysr2sympy("sqrt(2)"), np.sqrt(2)),
+            (sympy.exp(2), np.exp(2)),
+            (sympy.log(4), np.log(4)),
+            (sympy.sin(1), np.sin(1)),
+        ]
+
+        for expr, expected in test_cases:
+            m = pysr.export_torch.sympy2torch(expr, [])
+            result = m(self.torch.randn(10, 1))
+            np.testing.assert_almost_equal(result.item(), expected, decimal=3)
+
+        # Test with variables: sqrt(2) * x
+        x = sympy.symbols("x")
+        expr = sympy.sqrt(2) * x
+        m = pysr.export_torch.sympy2torch(expr, [x])
+        X = np.random.randn(10, 1)
+        np.testing.assert_almost_equal(
+            m(self.torch.tensor(X)).detach().numpy().flatten(),
+            np.sqrt(2) * X[:, 0],
+            decimal=3,
+        )
+
     def test_feature_selection_custom_operators(self):
         rstate = np.random.RandomState(0)
         X = pd.DataFrame({f"k{i}": rstate.randn(2000) for i in range(10, 21)})