Browse code
use pytest.mark.parametrize
Bas Nijholt authored on 17/03/2019 17:31:04
Showing 1 changed files
| ... | ... |
@@ -1,5 +1,7 @@ |
| 1 | 1 |
# -*- coding: utf-8 -*- |
| 2 | 2 |
|
| 3 |
+import pytest |
|
| 4 |
+ |
|
| 3 | 5 |
from adaptive.learner import Learner1D, BalancingLearner |
| 4 | 6 |
from adaptive.runner import simple |
| 5 | 7 |
|
| ... | ... |
@@ -24,32 +26,30 @@ def test_balancing_learner_loss_cache(): |
| 24 | 26 |
assert bl.loss(real=True) == real_loss |
| 25 | 27 |
|
| 26 | 28 |
|
| 27 |
-def test_distribute_first_points_over_learners(): |
|
| 28 |
- for strategy in ['loss', 'loss_improvements', 'npoints']: |
|
| 29 |
- learners = [Learner1D(lambda x: x, bounds=(-1, 1)) for i in range(10)] |
|
| 30 |
- learner = BalancingLearner(learners, strategy=strategy) |
|
| 31 |
- points, _ = learner.ask(100) |
|
| 32 |
- i_learner, xs = zip(*points) |
|
| 33 |
- # assert that are all learners in the suggested points |
|
| 34 |
- assert len(set(i_learner)) == len(learners) |
|
| 35 |
- |
|
| 36 |
- |
|
| 37 |
-def test_ask_0(): |
|
| 38 |
- for strategy in ['loss', 'loss_improvements', 'npoints']: |
|
| 39 |
- learners = [Learner1D(lambda x: x, bounds=(-1, 1)) for i in range(10)] |
|
| 40 |
- learner = BalancingLearner(learners, strategy=strategy) |
|
| 41 |
- points, _ = learner.ask(0) |
|
| 42 |
- assert len(points) == 0 |
|
| 43 |
- |
|
| 44 |
- |
|
| 45 |
-def test_strategies(): |
|
| 46 |
- goals = {
|
|
| 47 |
- 'loss': lambda l: l.loss() < 0.1, |
|
| 48 |
- 'loss_improvements': lambda l: l.loss() < 0.1, |
|
| 49 |
- 'npoints': lambda bl: all(l.npoints > 10 for l in bl.learners) |
|
| 50 |
- } |
|
| 51 |
- |
|
| 52 |
- for strategy, goal in goals.items(): |
|
| 53 |
- learners = [Learner1D(lambda x: x, bounds=(-1, 1)) for i in range(10)] |
|
| 54 |
- learner = BalancingLearner(learners, strategy=strategy) |
|
| 55 |
- simple(learner, goal=goal) |
|
| 29 |
+@pytest.mark.parametrize('strategy', ['loss', 'loss_improvements', 'npoints'])
|
|
| 30 |
+def test_distribute_first_points_over_learners(strategy): |
|
| 31 |
+ learners = [Learner1D(lambda x: x, bounds=(-1, 1)) for i in range(10)] |
|
| 32 |
+ learner = BalancingLearner(learners, strategy=strategy) |
|
| 33 |
+ points, _ = learner.ask(100) |
|
| 34 |
+ i_learner, xs = zip(*points) |
|
| 35 |
+ # assert that are all learners in the suggested points |
|
| 36 |
+ assert len(set(i_learner)) == len(learners) |
|
| 37 |
+ |
|
| 38 |
+ |
|
| 39 |
+@pytest.mark.parametrize('strategy', ['loss', 'loss_improvements', 'npoints'])
|
|
| 40 |
+def test_ask_0(strategy): |
|
| 41 |
+ learners = [Learner1D(lambda x: x, bounds=(-1, 1)) for i in range(10)] |
|
| 42 |
+ learner = BalancingLearner(learners, strategy=strategy) |
|
| 43 |
+ points, _ = learner.ask(0) |
|
| 44 |
+ assert len(points) == 0 |
|
| 45 |
+ |
|
| 46 |
+ |
|
| 47 |
+@pytest.mark.parametrize('strategy, goal', [
|
|
| 48 |
+ ('loss', lambda l: l.loss() < 0.1),
|
|
| 49 |
+ ('loss_improvements', lambda l: l.loss() < 0.1),
|
|
| 50 |
+ ('npoints', lambda bl: all(l.npoints > 10 for l in bl.learners)),
|
|
| 51 |
+]) |
|
| 52 |
+def test_strategies(strategy, goal): |
|
| 53 |
+ learners = [Learner1D(lambda x: x, bounds=(-1, 1)) for i in range(10)] |
|
| 54 |
+ learner = BalancingLearner(learners, strategy=strategy) |
|
| 55 |
+ simple(learner, goal=goal) |