From fbb460d4d0a5ed7d4222f8b3cb7f4d705bfaba74 Mon Sep 17 00:00:00 2001
From: Sait Cakmak <saitcakmak@meta.com>
Date: Tue, 27 Feb 2024 15:06:05 -0800
Subject: [PATCH] Patch codecov for SaasFullyBayesianSingleTaskGP &
 FullyBayesianPosterior

Summary: --

Reviewed By: Balandat, SebastianAment

Differential Revision: D54278602

fbshipit-source-id: 259fc73e03a68a13507b1574c50ca21fca5d86b2
---
 test/models/test_fully_bayesian.py | 26 +++++++++++++++++++++++++-
 1 file changed, 25 insertions(+), 1 deletion(-)

diff --git a/test/models/test_fully_bayesian.py b/test/models/test_fully_bayesian.py
index f0a4b7d1b6..86f4ad247f 100644
--- a/test/models/test_fully_bayesian.py
+++ b/test/models/test_fully_bayesian.py
@@ -50,7 +50,11 @@
     SaasPyroModel,
 )
 from botorch.models.transforms import Normalize, Standardize
-from botorch.posteriors.fully_bayesian import batched_bisect, GaussianMixturePosterior
+from botorch.posteriors.fully_bayesian import (
+    batched_bisect,
+    FullyBayesianPosterior,
+    GaussianMixturePosterior,
+)
 from botorch.sampling.get_sampler import get_sampler
 from botorch.utils.datasets import SupervisedDataset
 from botorch.utils.multi_objective.box_decompositions.non_dominated import (
@@ -752,6 +756,15 @@ def test_condition_on_observation(self):
                 torch.Size([num_models, num_train + num_cond, num_dims]),
             )
 
+            # With batch size only on Y.
+            cond_model = model.condition_on_observations(
+                cond_X_nobatch, cond_Y, noise=cond_Yvar
+            )
+            self.assertEqual(
+                cond_model.train_inputs[0].shape,
+                torch.Size([num_models, num_train + num_cond, num_dims]),
+            )
+
             # test repeated conditining
             repeat_cond_X = cond_X + 5
             repeat_cond_model = cond_model.condition_on_observations(
@@ -815,6 +828,17 @@ def f(x):
                     dist.cdf(x), q * torch.ones(1, 5, 1, **tkwargs), atol=1e-4
                 )
 
+    def test_deprecated_posterior(self) -> None:
+        mean = torch.randn(1, 5)
+        variance = torch.rand(1, 5)
+        covar = torch.diag_embed(variance)
+        mvn = MultivariateNormal(mean, to_linear_operator(covar))
+        with self.assertWarnsRegex(
+            DeprecationWarning, "`FullyBayesianPosterior` is marked for deprecation"
+        ):
+            posterior = FullyBayesianPosterior(distribution=mvn)
+        self.assertIsInstance(posterior, GaussianMixturePosterior)
+
 
 class TestPyroCatchNumericalErrors(BotorchTestCase):
     def tearDown(self):