[ase-users] PathInterpolator

Michael Joseph Waters michael.j.waters at northwestern.edu
Sat Feb 22 04:25:22 CET 2020


Hi ASE devs,

Following up on my comment to Reza, here is my path interpolator that uses scipy's interp1d to make very smooth atomic motion interpolation. You can do arbitrary refinement or use cubic splines for animations. The implementation isn't superb, but it is clean.

Here's what it looks like in use for a reaction pathway:

old_images = [  io.read('min1.CONTCAR'),
                io.read('saddle.CONTCAR'),
                io.read('min2.CONTCAR')]
num_new_images = 15
scaled_path_coordinates = np.linspace(0, 1.0, num_new_images )

my_interpolator = PathInterpolator(old_images)
new_images = my_interpolator(scaled_path_coordinates)

I think it might go nicely in geometry tools.

Cheers,
-Mike
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.fysik.dtu.dk/pipermail/ase-users/attachments/20200222/0f25ff48/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: min1.CONTCAR
Type: application/octet-stream
Size: 1656 bytes
Desc: min1.CONTCAR
URL: <http://listserv.fysik.dtu.dk/pipermail/ase-users/attachments/20200222/0f25ff48/attachment-0003.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: min2.CONTCAR
Type: application/octet-stream
Size: 1656 bytes
Desc: min2.CONTCAR
URL: <http://listserv.fysik.dtu.dk/pipermail/ase-users/attachments/20200222/0f25ff48/attachment-0004.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: path_interpolator.py
Type: text/x-python
Size: 4810 bytes
Desc: path_interpolator.py
URL: <http://listserv.fysik.dtu.dk/pipermail/ase-users/attachments/20200222/0f25ff48/attachment-0001.py>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: saddle.CONTCAR
Type: application/octet-stream
Size: 1656 bytes
Desc: saddle.CONTCAR
URL: <http://listserv.fysik.dtu.dk/pipermail/ase-users/attachments/20200222/0f25ff48/attachment-0005.obj>


More information about the ase-users mailing list