Iván Pulido

> How do we want to approach this? Just changing the move in this test would be enough to test what we want to test? From discussions with @zhang-ivy ,...

As discussed in our dev sync, I saw the following results  According to log output it would be taking more than 4 hours for completion, which seems like too...

~While debugging the performance of this test a solution to issue #1085 was required, I updated the description of this PR to reflect this.~ This was reverted, check discussion in...

@zhang-ivy Run the neutral transformation test with phenol to paracetamol, using the Vanilla HTF and the results for the DDG and dDDG are as follow: ``` n_iterations: 1000, DDG (4.379856483835027),...

> @ijpulidos : Did you run phenol to paracetamol with the rest htf as well? RestHTF gives better results for the DDG vs. dDDG comparison ``` n_iterations: 1000, DDG (0.7389286664047461),...

> I found a "strange" thing happening with the mapper, which may explain the differences. Left is RestHTF, right is Vanilla HTF I don't think this is a bug, it's...

For the `PointMutationExecutor` check https://github.com/choderalab/perses/pull/1030#discussion_r900267666 . This can be accomplish by using pytest fixtures and avoiding redundant code and efforts. In general, we want that in many of our tests.

We can parametrize the different combinations in `test_name` and `use_point_energies` for `run_unsampled_endstates_energies` in `test_unsampled_endstate_energies_GPU` test.

@SZhang66 thanks for the feedback! I cannot reproduce the errors you mention. I am directly running the example using `perses-cli --yaml setup_new.yaml` and it is indeed taking a long time...

I think for this one it makes sense to write our own `retry` decorator and use that in the functions that need it, such as the multisampler `equilibrate()` or `minimize()`...