add utility for printing shapes as strings (pytorch#1947)

Summary:
Pull Request resolved: pytorch#1947

This is useful for printing exception messages in a nice format that is easily verified with `assertRaisesRegex`.

Reviewed By: Balandat

Differential Revision: D47710862

fbshipit-source-id: 8f1a049fcd775889e086918ab8c9edf944fab8c6

botorch/logging.py

@@ -6,6 +6,7 @@
 
 import logging
 
+import torch
 
 LOG_LEVEL_DEFAULT = logging.CRITICAL
 
@@ -36,4 +37,8 @@ def _get_logger(
     return logger
 
 
+def shape_to_str(shape: torch.Size) -> str:
+    return f"`{' x '.join(str(i) for i in shape)}`"
+
+
 logger = _get_logger()

botorch/models/model.py

@@ -33,6 +33,7 @@
 import torch
 from botorch import settings
 from botorch.exceptions.errors import BotorchTensorDimensionError, InputDataError
+from botorch.logging import shape_to_str
 from botorch.models.utils.assorted import fantasize as fantasize_flag
 from botorch.posteriors import Posterior, PosteriorList
 from botorch.sampling.base import MCSampler
@@ -581,8 +582,8 @@ def fantasize(
             ):
                 raise BotorchTensorDimensionError(
                     f"Expected evaluation_mask of shape `{X.shape[0]} "
-                    f"x {self.num_outputs}`, but got `"
-                    f"{' x '.join(str(i) for i in evaluation_mask.shape)}`."
+                    f"x {self.num_outputs}`, but got "
+                    f"{shape_to_str(evaluation_mask.shape)}."
                 )
             if not isinstance(sampler, ListSampler):
                 raise ValueError("Decoupled fantasization requires a list of samplers.")

test/test_logging.py

@@ -6,8 +6,10 @@
 
 import logging
 
+import torch
+
 from botorch import settings
-from botorch.logging import LOG_LEVEL_DEFAULT, logger
+from botorch.logging import LOG_LEVEL_DEFAULT, logger, shape_to_str
 from botorch.utils.testing import BotorchTestCase
 
 
@@ -31,3 +33,9 @@ def test_settings_log_level(self):
             self.assertEqual(logger.level, logging.INFO)
         # Finally, verify the original level is set again
         self.assertEqual(logger.level, LOG_LEVEL_DEFAULT)
+
+    def test_shape_to_str(self):
+        self.assertEqual("``", shape_to_str(torch.Size([])))
+        self.assertEqual("`1`", shape_to_str(torch.Size([1])))
+        self.assertEqual("`1 x 2`", shape_to_str(torch.Size([1, 2])))
+        self.assertEqual("`1 x 2 x 3`", shape_to_str(torch.Size([1, 2, 3])))