| 28 Nov 2024 |
 Toma Susi | The former denote fractional coordinates, the latter operates on the actual positions | 00:00:54 |
| Toma Susi | For what it's worth, we use fractional = True as keyword in abTEM when specifying line profiles or area scans over the atoms object | 00:04:53 |
| Toma Susi | but there it's not the positions of atoms that are fractional, just the extents | 00:05:13 |
 mjwaters | In reply to @tomasusi:matrix.org
scaled_positions is a very different thing than scale_atoms, though Would say that the phonetic similarity is false friend for non-native English speakers? | 00:11:10 |
| Toma Susi | One might say that | 00:12:38 |
| Toma Susi | Though I would just argue that the similarity is rather a reason to revise the name than keep it | 00:13:02 |
| mjwaters | In reply to @tomasusi:matrix.org
Though I would just argue that the similarity is rather a reason to revise the name than keep it That's what I mean by false friend. | 00:21:56 |
| Toma Susi | Does ASE support this mmCIF format? I just heard from some folks it's becoming the standard in the bio TEM field.
https://pdb101.rcsb.org/learn/guide-to-understanding-pdb-data/beginner’s-guide-to-pdbx-mmcif | 05:11:41 |
| mjwaters | Interesting, I wonder if it is still headache-y to parse. | 05:55:12 |
| Toma Susi | Yeah I wonder, CIF is not great | 05:55:40 |
| mjwaters | Do you also get cifs from collaborators that are several megabytes and need to have absurd amounts for X-ray data stripped out before they can be parsed? | 05:57:39 |
| mjwaters | Toma Susi: I don't see anything in io that looks like mmCIF. | 06:02:16 |
| mjwaters | There's cif which links to the cif 1.1 spec: https://gitlab.com/ase/ase/-/blob/master/ase/io/cif.py?ref_type=heads | 06:02:37 |
| mjwaters | ase.io.formats.py has the list of supported formats. | 06:03:34 |
| mjwaters | https://gitlab.com/ase/ase/-/blob/master/ase/io/formats.py?ref_type=heads | 06:03:36 |
| mjwaters | Ask Hjorth Larsen: for formats that are output-only because they are not atomic structure files, like .png or .pov, how do tests work? We can make sure that the output matches a reference in someway right? i.e. we expect so many floats in certain sequence or that the data is the right length. How would we confirm that the output conforms to the standard? I ask because I've been talking to a graduate student about format that embeds base64 encoded arrays in .json files. There are io libraries out there, but it's not clear if it would be more or less work to use one. | 06:13:34 |
| mjwaters | Maybe we can talk tomorrow since it's a holiday here. | 06:13:59 |
 Ask Hjorth Larsen | Short answer is they probably don't, but we can discuss tomorrow | 19:28:51 |
| Ask Hjorth Larsen | I'd say that the test could check that certain structures are found in the output, that would be the overall goal. | 19:29:29 |
| 29 Nov 2024 |
| Toma Susi | Hmm let's see... I have a 3D array of 0/1 values as a mask that I want to apply to an Atoms object to remove atoms that fall into the 0 entries of the array based on their positions | 01:10:03 |
| Toma Susi | This should be fairly straightforward hmm | 01:10:17 |
| Toma Susi | but not quite trivial | 01:10:27 |
| Toma Susi | oh there is a cut() function | 01:11:00 |
| Toma Susi | but no that won't work, the volume I am cutting out has a complicated shape | 01:11:45 |
| Toma Susi | Ahh, got it | 01:57:09 |
| Toma Susi | fun, a 33-million atom "nano"particle | 06:05:48 |
| Toma Susi | ASE can just about handle that | 06:06:05 |
| Toma Susi | abTEM seems to not have too much trouble | 06:06:24 |
|  Kavinda Athapaththu joined the room. | 07:53:57 |
| 5 Dec 2024 |
| Ask Hjorth Larsen | Bar is open over in #ase-bar:matrix.org | 19:59:59 |
