A python module for solving the TISE using discrete-variable representation.
- Solve ND problems using sine-DVR and CM-DVR
- Solve ND problems using DVR type algorithms learned from program synthesis (
synthesised_algorithms.py). - Contains the AMMPOT4 NH3 and Partridge-Schwenke H2O potential.
- Contains routines for generating grids in normal and cartesian coordinates, as well as transforming between.
- Contains routines for diagonalising the position operator to construct a DVR (HEG procedure for diagonalisation DVR).
- Generate uniform direct product grids, or quasi-random Sobol grids for use in training ML potentials on fewer points.
Install requires numpy ~= 1.26 for f2py compilation of FORTRAN routines.
To install:
> git clone https://github.com/KyleAcheson/NDDVR.git
> cd NDDVR
> pip install .
This will install all source files in the default location for your setup as /path/fast_dvr/, using venvs is recommended.
Import as:
import fast_dvr