| 21 Nov 2024 |
 mjwaters | yes, you have to. | 20:01:55 |
| mjwaters | those functions in geometry https://gitlab.com/ase/ase/-/blob/master/ase/geometry/geometry.py?ref_type=heads#L251 | 20:02:15 |
| mjwaters | they don't accept the atoms object so they have no way of knowing the periodicity otherwise. | 20:02:46 |
 askhl | navidmrch: I don't know much about the Analysis class, but I'd guess it respects mic/PBCs since it has the atoms inside and uses a neighbor list internally. The standalone functions need mic/PBCs explicitly as mjwaters says | 20:06:29 |
 navidmrch | In reply to @mjwaters:matrix.org
they don't accept the atoms object so they have no way of knowing the periodicity otherwise. mjwaters Ask Hjorth Larsen Thank you for the responses. You are right. So, for functions that use atoms object like 'neighbor_list', there is no need to use 'mic/PBCs' option. | 20:08:54 |
| mjwaters | if you use atoms.get_angles() it will use the PBCs in your system | 20:09:40 |
| mjwaters | https://gitlab.com/ase/ase/-/blob/master/ase/atoms.py?ref_type=heads#L1713 | 20:09:43 |
| askhl | navidmrch: Indeed. However it is always wise to be careful and verify that things work as you expect when using a new feature | 20:10:10 |
| mjwaters | In reply to @askhl:matrix.org
navidmrch: Indeed. However it is always wise to be careful and verify that things work as you expect when using a new feature And we should make sure that we have sensible defaults. Because you develop and I promote ASE :) | 20:11:40 |
| mjwaters | The Analysis class seems weirdly constructed, I'm not sure what the goal is for it. | 20:12:31 |
 Toma Susi | In reply to @mjwaters:matrix.org
What do you mean? do you want the polygon mesh and vertices ? or have a point cloud programmatically from the symmetry of the shape? I want an ASE Atoms object nanoparticle of that shape | 22:24:45 |
| mjwaters | oooooooh. that makes sense. | 22:30:31 |
| mjwaters | I think ASE can do all the wulf shapes for simple lattices. | 22:31:32 |
| mjwaters | if it's not in there, then it gets a bit harder. | 22:32:11 |
| Toma Susi | Hmm right Wulff shapes | 22:40:44 |
| Toma Susi | This would be for FCC and fairly standard facets | 22:40:54 |
| Toma Susi | I'll look into that class | 22:41:05 |
| Toma Susi | Is it just a matter of trying different surface energies until I get what I want? | 22:41:46 |
| Toma Susi | Ooh WulffPack does twinned boundaries, which is what I also need:
https://wulffpack.materialsmodeling.org/examples/basic_usage.html | 22:43:05 |
| Toma Susi | Paul Erhart again, I knew it from the logo! | 22:43:45 |
| 22 Nov 2024 |
 tgmaxson | Only thing missing from WulffPack is Winterbottom structures <90 degree contact angle (the interfacial energy should be negative in this case and it does not have any logic for this). | 08:24:09 |
| 23 Nov 2024 |
| mjwaters | Wait, has the MACE-MP0 paper not been published by a journal? | 05:10:25 |
 Adam Jackson | Seems so! They are onto third iteration already with MACE-MP-0b2. (Adding a character with each version? Looking forward to MACE-MP-0b2iii-FINAL-FINAL 😆) | 08:01:34 |
| mjwaters | -No-but-really-final-v2 | 08:04:44 |
| 27 Nov 2024 |
 jjmo | ASE uses the term "scaled positions" for positions in units of the lattice vectors. Is there a better term? "Relative"? | 16:07:12 |
 Robert Warmbier | Relative sounds good. Alternatively, crystal positions. But relative sounds better, me think | 16:11:45 |
| Adam Jackson | There's a MR for "fractional coordinates" https://gitlab.com/ase/ase/-/merge_requests/1537 | 17:09:32 |
| mjwaters | I don't love that I'd have to update a bunch of my scripts and code examples I've made for myself and I like that scaled_positions has phonetic similarity to the scale_atoms option for atoms.set_cell(). But obvious variable names are important. | 17:48:40 |
| Adam Jackson | Clearly this is a deprecation that would impact pretty-much everyone and so should be handled slowly and carefully. | 18:02:43 |
| 28 Nov 2024 |
| Toma Susi | scaled_positions is a very different thing than scale_atoms, though | 00:00:37 |
