fix: Cleanup relax_optim

examples/8_trapping_sindy_examples/trapping_sindy_paper_examples.py

@@ -520,9 +520,6 @@
 alpha_m = 5.0e-1 * eta
 
 
-# run trapping SINDy with "relax_optim = False" here, uses CVXPY now
-# so this tends to be much slower but often need far fewer algorithm iterations.
-# For this problem, a single (very slow) update is all that is needed!
 sindy_opt = ps.TrappingSR3(
     _n_tgts=6,
     _include_bias=True,
@@ -594,7 +591,6 @@
 #
 # $$
 # \begin{align}
-# \label{eq:burgers}
 #     \dot{u} &= -(U + u)\partial_x u + \nu \partial_{xx}^2u + g(x,t),
 # \end{align}
 # $$
@@ -605,7 +601,6 @@
 #
 # $$
 # \begin{equation}
-# \label{eq:burgers_galerkin}
 #     \dot{a}_k = \left( \delta_{|k|1} \sigma - \nu k^2  - ikU \right) a_k - \sum_{\ell=-r}^{r} i \ell a_{\ell} a_{k - \ell}.
 # \end{equation}
 # $$

pysindy/optimizers/trapping_sr3.py

@@ -201,6 +201,9 @@ def __init__(
         A0: Union[NDArray, None] = None,
         **kwargs,
     ):
+        self.alpha = alpha
+        self.beta = beta
+        self.mod_matrix = mod_matrix
         # n_tgts, constraints, etc are data-dependent parameters and belong in
         # _reduce/fit ().  The following is a hack until we refactor how
         # constraints are applied in ConstrainedSR3 and MIOSR
@@ -285,10 +288,8 @@ def __post_init_guard(self):
             raise ValueError("gamma must be negative")
         if self.tol <= 0 or self.tol_m <= 0 or self.eps_solver <= 0:
             raise ValueError("tol and tol_m must be positive")
-        if self.inequality_constraints and self.relax_optim and self.threshold == 0.0:
-            raise ValueError(
-                "Ineq. constr. -> threshold!=0 + relax_optim=True or relax_optim=False."
-            )
+        if self.inequality_constraints and self.threshold == 0.0:
+            raise ValueError("Inequality constraints requires threshold!=0")
 
     def set_params(self, **kwargs):
         super().set_params(**kwargs)

test/test_optimizers.py

@@ -510,7 +510,6 @@ def test_stable_linear_sr3_linear_library():
     [
         dict(),
         dict(accel=True),
-        dict(relax_optim=False),
     ],
 )
 @pytest.mark.parametrize(
@@ -938,7 +937,7 @@ def test_constrained_inequality_constraints(data_lorenz, params):
 )
 def test_trapping_cost_function(params):
     expected = params.pop("expected")
-    opt = TrappingSR3(relax_optim=True, **params)
+    opt = TrappingSR3(**params)
     x = np.eye(2)
     y = np.ones(2)
     xi, cost = opt._create_var_and_part_cost(2, x, y)
@@ -960,17 +959,6 @@ def test_trapping_inequality_constraints():
         constraint_rhs=constraint_rhs,
         constraint_order="feature",
         inequality_constraints=True,
-        relax_optim=True,
-    )
-    opt.fit(x, y)
-    assert np.all(np.dot(constraint_matrix, (opt.coef_).flatten()) <= constraint_rhs)
-    # Run Trapping SR3 with CVXPY for the m solve
-    opt = TrappingSR3(
-        constraint_lhs=constraint_matrix,
-        constraint_rhs=constraint_rhs,
-        constraint_order="feature",
-        inequality_constraints=True,
-        relax_optim=False,
     )
     opt.fit(x, y)
     assert np.all(np.dot(constraint_matrix, (opt.coef_).flatten()) <= constraint_rhs)
@@ -1023,7 +1011,6 @@ def test_inequality_constraints_reqs():
             constraint_rhs=constraint_rhs,
             constraint_order="feature",
             inequality_constraints=True,
-            relax_optim=True,
         )