add finished property to manual generators (#409)

Summary:
Pull Request resolved: #409

This adds a finished property to the manual generator classes so that they can keep track of whether they have generated all their points. Once this is hooked into the strategy's finishing logic, it should make writing configs simpler.

Reviewed By: JasonKChow

Differential Revision: D64600239

fbshipit-source-id: f8d2c18f516592a09cd545e76b52a8878b22b66b

Author: crasanders

aepsych/generators/manual_generator.py

@@ -6,15 +6,16 @@
 # LICENSE file in the root directory of this source tree.
 
 import warnings
-from typing import Optional, Union, Dict
+from typing import Dict, Optional, Union
 
 import numpy as np
 import torch
+from torch.quasirandom import SobolEngine
+
 from aepsych.config import Config
 from aepsych.generators.base import AEPsychGenerator
 from aepsych.models.base import AEPsychMixin
 from aepsych.utils import _process_bounds
-from torch.quasirandom import SobolEngine
 
 
 class ManualGenerator(AEPsychGenerator):
@@ -95,6 +96,10 @@ def get_config_options(cls, config: Config, name: Optional[str] = None) -> Dict:
 
         return options
 
+    @property
+    def finished(self):
+        return self._idx >= len(self.points)
+
 
 class SampleAroundPointsGenerator(ManualGenerator):
     """Generator that samples in a window around reference points in a predefined list."""
@@ -131,9 +136,9 @@ def __init__(
             grid = self.engine.draw(samples_per_point)
             grid = p_lb + (p_ub - p_lb) * grid
             generated.append(grid)
-        generated = torch.Tensor(np.vstack(generated)) #type: ignore
+        generated = torch.Tensor(np.vstack(generated))  # type: ignore
 
-        super().__init__(lb, ub, generated, dim, shuffle, seed) #type: ignore
+        super().__init__(lb, ub, generated, dim, shuffle, seed)  # type: ignore
 
     @classmethod
     def get_config_options(cls, config: Config, name: Optional[str] = None) -> Dict:

tests/generators/test_manual_generator.py

@@ -9,6 +9,7 @@
 
 import numpy as np
 import numpy.testing as npt
+
 from aepsych.config import Config
 from aepsych.generators import ManualGenerator, SampleAroundPointsGenerator
 
@@ -50,6 +51,7 @@ def test_manual_generator(self):
         gen = ManualGenerator.from_config(config)
         npt.assert_equal(gen.lb.numpy(), np.array([0, 0]))
         npt.assert_equal(gen.ub.numpy(), np.array([1, 1]))
+        self.assertFalse(gen.finished)
 
         p1 = list(gen.gen()[0])
         p2 = list(gen.gen()[0])
@@ -60,6 +62,7 @@ def test_manual_generator(self):
         self.assertEqual(sorted([p1, p2, p3, p4]), points)
         self.assertEqual(gen.max_asks, len(points))
         self.assertEqual(gen.seed, 123)
+        self.assertTrue(gen.finished)
 
 
 class TestSampleAroundPointsGenerator(unittest.TestCase):
@@ -86,13 +89,16 @@ def test_sample_around_points_generator(self):
         npt.assert_equal(gen.ub.numpy(), np.array([1, 1]))
         self.assertEqual(gen.max_asks, len(points * samples_per_point))
         self.assertEqual(gen.seed, 123)
+        self.assertFalse(gen.finished)
 
         points = gen.gen(gen.max_asks)
         for i in range(len(window)):
             npt.assert_array_less(points[:, i], points[:, i] + window[i])
             npt.assert_array_less(np.array([0] * len(points)), points[:, i])
             npt.assert_array_less(points[:, i], np.array([1] * len(points)))
 
+        self.assertTrue(gen.finished)
+
 
 if __name__ == "__main__":
     unittest.main()