-
Notifications
You must be signed in to change notification settings - Fork 23
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
🐛 21 Physically valid layouts (for the same layout) not detected by QuickExact
#385
Comments
Thank you for this interesting discovery! 🙏 |
Here's a more minimal example, and interestingly enough, QuickExact find one physically valid layout with a positive charge, though I have another physically valid layout that it didn't find that does not have a positive charge. One to grind the gears on for sure, though I bet 3 DBs is a lot more managable to work with :) This one operates on all default parameters much like the former.
|
I should add: ExGS does not give the same issue (as one would hope) |
Thank you for your investigation! I think I have a strong feeling where the bug is. |
I found another issue with QuickExact results. (do I open a different issue for this one?) The test case below shows evidence of QuickExact returning duplicate results for this layout. I haven't been able to find a more minimal example thus far unfortunately.
|
Thank you so much! |
QuickExact
fiction version
fiction main -- latest to date (62a132f)
OS
Ubuntu 22.04.2 LTS | Linux 5.15.133.1-microsoft-standard-WSL2 x86_64 GNU/Linux
Python version
3.11.2
C++ compiler
Ubuntu clang version 14.0.0-1ubuntu1.1
Additional environment information
No response
Description
There's not much more to say than the title states. I found a 20 DB layout for which QuickExact gives 0 physically valid layouts, while the 21 charge configurations in the "how to reproduce" section are actually physically valid as judged by
charge_distribution_surface
standard functionality. I suspect the presence of positive charges here brings up an issue. I'm very much curious to hear what the problem here is when the issue is discovered.To be sure there is no issue with (physical) parameters, I have specified them explicitly, although the issue is reproduced when all parameters (of
charge_distribution_surface
andquickexact
) are defaulted.You will have to take my word for that the charge configurations below are all unique (either that or check it yourself I suppose). Though even if they weren't and only one physically valid charge distribution is not detected by QuickExact, the issue claim here would still be valid.
Expected behavior
QuickExact should find all physically valid layouts, hence also these ones. Either that or these layouts should not be judged physically valid, but I suspect the former is more likely.
How to Reproduce
Insert the following Catch2 test into a test file of choice and run it. On my systems the test passes, meaning all of these 21 charge distributions are physically valid, and QuickExact finds precisely 0 physically valid charge distributions for the same layout.
The text was updated successfully, but these errors were encountered: