[ase-users] ASE Development Proposal

David Ollodart davidollodart at gmail.com
Sun Nov 17 19:47:44 CET 2019


Per the instructions in `ase/development/contribute.html` I am sending a
message to the mailing list regarding what I propose to develop. I would
like to develop those things I have posted on
`ase/ase-workshop-discussion/issues`. But I have a few straightforward
tasks which can be accomplished shortly:

1. Add a dropdown menu to the toolbar which will display in the status line
center of mass, average position (which is different if the center of mass
is actually mass). Since some users may not want their toolbar cluttered,
add options in `ase/gui/defaults.py` to select which toolbar menus are
enabled.

Will it be decided how to go forward with the GUI (Tkinter, vtk as in the
user list serv message 005030, or other) in the ase workshop?
2. Add `Atoms` methods to calculate all bond angles and all dihedral angles
bond angle analogous to `get_all_distances()`. This would use the
connectivity matrix from `ase.neighborlist` and array masks to limit the
triples and quadruples for which angles and dihedral angles are calculated,
e.g., to only sets of atoms each having at least one bond to one other, or
triples of atoms containing 2 of one element and 1 of another.
3. Add geometry optimization taking into account only volume exclusion for
confined molecules, e.g., in zeolites and MOFs. This can be done by
optimization using discontinuous penalty functions based on covalent radii
overlap. For the example problem of zeolite confinement there is the center
of mass of the molecule fixed in the center of the zeolite void and random
changes in the orientation can be sampled in the azimuth and zenith angles.
If not a part of the package this may be an example for the documentation.
4. Improve the `sort` function in `ase.build.utils`. Include hierarchical
sorting based on Cartesian and crystal coordinates, sorting along any
miller index using the `ase.geometry.get_layers` function, and sorting by
distance from a point, a line, or a plane. This may also just be an example
or set of examples for the documentation.
5. Add subcommand `diff` to the ase command line interface to diff
quantities like atom positions, forces on atoms, total energy between two
coordinate or calculation output files. This would provide the same
functionality but with general I/O as that from the `vtsttools` scripts
developed for VASP by the Henkelmann group.

I have reviewed the issue trackers and am interested in working on
`ase-workshop-dissusion/issues/#9` and `ase/issues/#468`. I welcome
comments and suggestions.

Best,

David Ollodart
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.fysik.dtu.dk/pipermail/ase-users/attachments/20191117/d4ba2025/attachment.html>


More information about the ase-users mailing list