1: \begin{abstract}

2: We propose an efficient, accurate method to integrate the basins of

3: attraction of a smooth function defined on a general discrete grid, and

4: apply it to the Bader charge partitioning for the electron charge density.

5: Starting with the evolution of trajectories in space following the gradient

6: of charge density, we derive an expression for the fraction of space

7: neighboring each grid point that flows to its neighbors.  This serves as

8: the basis to compute the fraction of each grid volume that belongs to a

9: basin (Bader volume), and as a weight for the discrete integration of

10: functions over the Bader volume.  Compared with other grid-based

11: algorithms, our approach is robust, more computationally efficient with

12: linear computational effort, accurate, and has quadratic convergence.

13: Moreover, it is straightforward to extend to non-uniform grids, such as

14: from a mesh-refinement approach, and can be used to both identify basins of

15: attraction of fixed points and integrate functions over the basins.

16: \end{abstract}

17: