Oliver Beckstein

As discussed in #208 , alchemlyb 2.x would depend on pymbar ≥ 4.0

I've been pretty busy all around. I'd like to make a 0.4.0 release #113 with all the good stuff and then we can revisit workflows.

@harlor do you want to continue the broader discussion in https://github.com/alchemistry/alchemlyb/discussions/159 — especially the more general part on structuring the library?

@xiki-tempula I am swamped this week and don't have time for big PR reviews. Can you please ping me again Monday next week? Sorry. Perhaps someone else has time to...

Yes, of course. Sorry, I am currently snowed under deadlines and it is has been difficult to commit much time to anything else.

That's an interesting idea. How will a user use the class? Can you show mock-code that shows how it would be used in practice?

This looks quite clean (and I obviously like the AnalysisBase approach from MDAnalysis.analysis ;-)). Presumably, we would have a `workflows.base.AnalysisBase` class and then derive, e.g., ```python class ABFEAnalysis(base.AnalysisBase): ... ```...

@xiki-tempula I won’t be able to look at it over the next week, please be patient. Thanks.

Tests fail with > ERROR collecting src/alchemlyb/tests/test_preprocessing.py Do you know why?

Yes, let’s do “false” to run all CI.