QCQMC Part 5: Add Trial Wavefunction #347
Conversation
fdmalone
force-pushed
the
qcqmc-5
branch
2 times, most recently
from
7c82c04 to
06435f4
Compare
fdmalone
changed the title
|
@mpharrigan / @dstrain115 PTAL |
recirq/qcqmc/trial_wf.py
Outdated
| n_optimization_restarts: int = 1, | ||
| do_print: bool = True, | ||
| use_fast_gradients: bool = False, | ||
| ) -> Tuple[np.ndarray, np.ndarray, np.ndarray]: |
There was a problem hiding this comment.
This is also a long and complicated function with no docstring or comments.
There was a problem hiding this comment.
Split it up and added a reference to the paper.
recirq/qcqmc/trial_wf.py
Outdated
| print("Two Body Rotation Parameters:") | ||
| print(two_body_params) | ||
|
|
||
| return one_body_params, two_body_params, one_body_basis_change_mat |
There was a problem hiding this comment.
I agree. I think this should be split up.
There was a problem hiding this comment.
Split everything into modules
|
What were you thinking re: splitting it up?
would be one way |
|
Pretty much this yes:
|
|
optimization might be split too if possible |
|
One thing the refactor revealed is a lack of unit tests for quite a few things... |
recirq/qcqmc/optimize_wf_test.py
Outdated
| assert trial_wf.ansatz_energy < -1.947 | ||
|
|
||
|
|
||
| # TODO: Speed up this test and add a similar one with non-trivial heuristic layers. |
There was a problem hiding this comment.
Maybe use @pytest.mark.slow on the relevant tests?
There was a problem hiding this comment.
@wjhuggins, do you remember where this is from? Running these tests only takes 7 seconds total which isn't necessarily slow, but I do see quite significant creep in the CI's duration (from 5m to 13! minutes) since these PRs started to go in. I'll follow up with a PR to make things faster.
There was a problem hiding this comment.
yeah somehow these are exceptionally slow for actions, but only take a few seconds on my laptop. I'll investigate
|
@dstrain115 could we merge and I'll follow up to add some more unit tests and fix the CI slowness. |
|
Yeah, that's fine with me. |
|
(I can't merge this myself) |
Add trial wavefunction module.
This is very big and could be split up into several files. WDYT @mpharrigan