Seth Axen
Seth Axen
@devmotion thanks for the reminder; this slipped off the end of my to-do list. I'll prioritize finishing this in the next few days. I'll let you know when it's ready...
@devmotion this should be ready for review now.
> So it seems in the end it is not necessary to change `zero(::NotImplemented)` and `zero(::Type{
Looking at the rules in `ChainRules`, we do use the tangents and cotangents differently. For the pushforwards, we always (I think) maintain the right type constraints and such so that...
We tried to add a timeout feature when `e3fp` was first released, but at the time, there wasn't a good way to do this in Python. It's worth revisiting to...
Thanks for the issue, and I apologize for the delayed reply. This looks like it might be due to a change in `rdkit.Chem.SDWriter`. I'll look into a workaround.
@mjke yes, but the `Fingerprinter` should also anticipate cases like this, fail gracefully with an informative warning, and proceed. We previously had issues when there were floating (unbound) molecules (see...
@8li to standardize, you'll need to `pip install standardiser` and then pass the `--standardise` parameter during *conformer generation*.
Commit 7ffc068 should prevent catastrophic failure in cases like this and all cases where an error occurs in fingerprinting. Still, we should anticipate this case and log an informative error...