new torontonian sampling #248
Conversation
…ction which includes displacement@
Codecov Report
@@ Coverage Diff @@
## master #248 +/- ##
=========================================
Coverage 100.00% 100.00%
=========================================
Files 21 21
Lines 1263 1260 -3
=========================================
- Hits 1263 1260 -3
Continue to review full report at Codecov.
|
thewalrus/tests/test_samples.py
Outdated
| @@ -429,6 +453,7 @@ def test_thermal_state_torontonian(self, sample_func): | |||
| @pytest.mark.parametrize("parallel", [True, False]) | |||
| def test_torontonian_sample_graph(self, parallel): | |||
| """Test torontonian sampling from a graph""" | |||
| np.random.seed(42) | |||
There was a problem hiding this comment.
How come this is needed here? I noticed that there's also a seed function imported on line 28, that basically does the exact same thing. Just for consistency maybe that could be used here as well.
There was a problem hiding this comment.
We put it in whilst debugging some things. Tests were passing on my PC but not in the CI. I'll try to see if I can get the CI tests to pass without the seed.
There was a problem hiding this comment.
okay, tests seem to pass without the seed! :D
Context:
the old torontonian sampling function using tor, and so could not be used for states with displacement. It was also a little clumsy.
Description of the Change:
The function now calls
threshold_detection_prob, making the code a bit cleaner and importantly also allows for displaced Gaussian states to be sampled from.Benefits:
Cleaner code, nonzero displacement now allowed.
Possible Drawbacks:
We could do some clever things to speed it up slightly.
Related GitHub Issues: