removed vestigial apply_new_noise member functions.

LLNL · Sep 27, 2023 · b328140 · b328140
1 parent 398e057
commit b328140
Show file tree

Hide file tree

Showing 2 changed files with 0 additions and 44 deletions.
diff --git a/MuyGPyS/gp/multivariate_muygps.py b/MuyGPyS/gp/multivariate_muygps.py
@@ -378,25 +378,3 @@ def optimize_scale(
             scales = mm.assign(scales, new_scale_val[0], i)
         self.scale._set(scales)
         return self
-
-    def apply_new_noise(self, new_noise):
-        """
-        Updates the heteroscedastic noise parameters of a MultivariateMuyGPs
-        model.
-
-        Args:
-            new_noise:
-                A matrix of shape
-                `(test_count, nn_count, nn_count, response_count)` containing
-                the measurement noise corresponding to the nearest neighbors
-                of each test point and each response.
-        Returns:
-            A MultivariateMuyGPs model with updated heteroscedastic noise
-            parameters.
-        """
-        ret = deepcopy(self)
-        for i, model in enumerate(ret.models):
-            model.noise = HeteroscedasticNoise(new_noise[:, :, :, i], "fixed")
-            model._mean_fn = PosteriorMean(model.noise)
-            model._var_fn = PosteriorVariance(model.noise, model.scale)
-        return ret
diff --git a/MuyGPyS/gp/muygps.py b/MuyGPyS/gp/muygps.py
@@ -337,28 +337,6 @@ def fast_posterior_mean(
         """
         return self._fast_posterior_mean_fn(Kcross, coeffs_tensor)
 
-    def apply_new_noise(self, new_noise: NoiseFn):
-        """
-        Updates the homo/heteroscedastic noise parameter(s) of a MuyGPs model.
-        To be used when the MuyGPs model has been trained and needs to be
-        used for prediction, or if multiple batches are needed during training
-        of a heteroscedastic model.
-
-        Args:
-            new_noise:
-                If homoscedastic, a float to update the nugget parameter.
-                If heteroscedastic, a matrix of shape
-                `(test_count, nn_count)` containing the measurement
-                noise corresponding to the nearest neighbors of each test point.
-        Returns:
-            A MuyGPs model with updated noise parameter(s).
-        """
-        ret = deepcopy(self)
-        ret.noise = new_noise
-        ret._mean_fn = PosteriorMean(ret.noise)
-        ret._var_fn = PosteriorVariance(ret.noise, ret.scale)
-        return ret
-
     def get_opt_mean_fn(self) -> Callable:
         """
         Return a posterior mean function for use in optimization.