Jason Swails

Ah, here were the changes I made way back when: https://github.com/drivendataorg/erdantic/compare/main...swails:erdantic:feature/plantuml-diagrams It looks like the code base has changed substantially since I did that, but this worked for me.

I think we should initialize a logger by default. loggers from the `logging` module are infinitely better than `print` for libraries you intend on including. It makes it trivially easy...

Why not just get rid of that package on pypi and then push one when it's ready?

If you're trying to find the absolute PPI free energies, the reason you're not seeing any tutorials is because this is an insanely hard problem. Because proteins are generally quite...

A few observations here. An OpenMM `System` object has a fixed number of particles, so you would need to use a separate `System` (and corresponding `Context`) object for each particle...

There is no tutorial for doing this explicitly. You can look at the MMPBSA.py script available from AmberTools to do this (and ParmEd can be used to generate the Amber...

If you install from conda using the following environment.yaml file: ```yaml name: xformers-test channels: - https://conda.anaconda.org/pytorch - https://conda.anaconda.org/xformers - https://conda.anaconda.org/dglteam/label/cu117 - https://conda.anaconda.org/nvidia - https://conda.anaconda.org/conda-forge dependencies: - python>=3.10 - pip -...

That's the name of the conda environment that gets created. The package is `xformers==0.25.0` (see the `dependencies` section of the YAML file). If you run `conda env create -f environment.yaml`...

I definitely understand the reluctance to support countless different flavors of a library to accommodate arbitrary environments. If this is indeed the decision (and I provide my meaningless support for...

You are coming at this from the perspective of an Amber user using one-off scripts -- I'd argue that the target audience of ambermini are _non_-Amber users. The purpose is...