Vanishing velocity about scikit-fmm HOT 3 CLOSED

Hi thanks for getting in touch. Can you be more specific about what you mean by vanishing velocity?

The best reference for fast marching methods is "Level Set Methods and Fast Marching Methods
Evolving Interfaces in Computational Geometry, Fluid Mechanics, Computer Vision, and Materials Science" J.A. Sethian, Cambridge University Press, 1999 Cambridge Monograph on Applied and Computational Mathematics

There is no formal publication about scikit-fmm, if you want to acknowledge it just write that you use the scikit-fmm Python module in you publication.

from scikit-fmm.

Hi @jkfurtney thanks for your reply. There were mask when setting phi, which I thought similar to setting velocity of those masked cell to zero. Could you explain what is happening when these masked cells for phi?

from scikit-fmm.

The input array phi can be a masked array. In the masked regions the front is not allowed to propagate. In the travel_time function, there is a numerical precision issue for very small values of velocity. So if velocity is less that numeric_limits::epsilon() (which I think is 2.220446049250313e-16 on most modern machines) the velocity is taken as zero and that region is masked off.

If that is confusing it would be great if you could clarify the documentation and make a pull request.

from scikit-fmm.