Multiwfn forum / Topological analysis on grid without interpolation

Re: Topological analysis on grid without interpolation

dummy@example.com (Alexey) — Mon, 13 Oct 2025 19:18:29 +0000

Thanks, Tian Lu

Re: Topological analysis on grid without interpolation

dummy@example.com (sobereva) — Mon, 13 Oct 2025 18:40:53 +0000

I have checked, at high-quality grid (corresponding to 0.13 Bohr grid spacing of your system), the 3D-spline interpolated density around some nuclei doesn't vary fully reasonably, making generation of a few bond paths unsuccessful. However, if grid spacing of 0.05 Bohr is used to generate electron density grid data, all bond paths can be successfully generated.

Unlike basin analysis in Multiwfn, the bond paths are not generated directly at grids. To generate bond paths, the derivative of electron density beyond grids must be evaluated, so it is impossible to "simply calculate the derivatives numerically on the grid"

Topological analysis on grid without interpolation

dummy@example.com (Alexey) — Sun, 12 Oct 2025 20:10:32 +0000

Hello, I've discovered that the topological analysis based on a .cub file doesn't find all the bond paths (I'm using -3 BSpline interpolation). Therefore, I'd like to know if it's possible to perform a topological analysis based on a .cub file without interpolation, i.e., simply calculate the derivatives numerically on the grid?

I've attached an example where I couldn't find the bond paths between atoms 3 and 5 (water dimer) using
"High-quality grid, covering the entire system, about 1,728,000 points in total"
01_Water-Water_3_5.wfn