GSOC 2025: Add electron density integration tools and BFit data processor #279
Conversation
- BFit data processor for s-type orbital data - UniformGrid vs CubicProTransform comparison test - Grid size performance analysis for 5 molecular systems
There was a problem hiding this comment.
Hi @tutou2356, this is alot better, and exactly finishes the last bit of Phase 1! Thank you for starting a fresh PR, cleaning up the files, and removing duplicates. Please before our next meeting, address the PR comments that I have written, and be prepared to showcase some results and explain the decisions you made surround the code. This will help us with our midterm evaluations. As always, if you need any help or questions, feel free to send an email or put on private discussions.
| return None, f"UniformGrid error: {str(e)}" | ||
|
|
||
|
|
||
| def create_promol_params(bfit_processor, symbols, coords, deps): |
There was a problem hiding this comment.
This is good function to have, and I would make it part of BFitDataProcessor, as a function there. We can think of BFitDataProcessor to be the middle object between BFit and PromolParams.
Add grid construction time and integration time columns to uniform vs transform comparison table, simplify duplicate code in test_grid_sizes function
This commit introduces Adaptive1DGrid, a class for adaptive numerical integration on the [-1, 1] interval. The implementation uses a segment-based strategy, where the grid refines locally by bisecting the segment with the largest integral contribution. Each segment is built upon a Gauss-Legendre quadrature rule.
tutou2356
force-pushed
the
electron-density-tools
branch
from
cf31254 to
aa7e243
Compare
Adds the AdaptiveUniformGrid wrapper class to provide recursive subdivision of a uniform grid.
- Rename bfit_data_processor.py to utils_promol.py and move to src/grid/ - Rename electron_density_integration.py to test_CubicProTransform.py and move to src/grid/tests - Add PromolMolecule class that integrates BFit data with CubicProTransform
-Filter tiny coefficients for numerical stability -Added a Jupyter Notebook that serves as an overview of the project
Ali-Tehrani
changed the title
2 participants
Files Added
s-type BFit data processor (
scripts/bfit_data_processor.py)UniformGrid vs CubicProTransform comparison test(
scripts/electron_density_integration.py)