Allow monotonic rejection GP to initialize its inducing point allocat…

…or from config. (#485)

Summary:
Pull Request resolved: #485

Monotonic rejection GP used to completely ignore any inducing point config, now it follows other models.

Differential Revision: D67225873

aepsych/models/monotonic_rejection_gp.py

@@ -355,15 +355,6 @@ def from_config(cls, config: Config) -> MonotonicRejectionGP:
         ub = config.gettensor(classname, "ub")
         dim = config.getint(classname, "dim", fallback=None)
 
-        inducing_point_method_class = config.getobj(
-            classname, "inducing_point_method", fallback=AutoAllocator
-        )
-        # Check if allocator class has a `from_config` method
-        if hasattr(inducing_point_method_class, "from_config"):
-            inducing_point_method = inducing_point_method_class.from_config(config)
-        else:
-            inducing_point_method = inducing_point_method_class()
-
         mean_covar_factory = config.getobj(
             classname, "mean_covar_factory", fallback=monotonic_mean_covar_factory
         )
@@ -381,7 +372,6 @@ def from_config(cls, config: Config) -> MonotonicRejectionGP:
             lb=lb,
             ub=ub,
             dim=dim,
-            inducing_point_method=inducing_point_method,
             num_induc=num_induc,
             num_samples=num_samples,
             num_rejection_samples=num_rejection_samples,