v1.1.9

Full Changelog: v1.1.8...v1.1.9

v1.1.8

Full Changelog: v1.1.7...v1.1.8

v1.1.7

What's Changed

• remove duplicate O2 by @jcwang587 in #25
• Update CITATION.cff and README citation to published Chem. Mater. article by @idocx in #27

New Contributors

• @jcwang587 made their first contribution in #25

Full Changelog: v1.1.6...v1.1.7

v.1.1.6

Downgrade Ray's version for compatibility.

v1.1.5

Fix the additional_cifs in Dara server api endpoint.

v1.1.4

Add a new parameter enable_angular_cut to control whether to run the search only in a part of the pattern to save some time. By default it is set to True (same behavior as before) but the user can disable it to run the search in full pattern, which is useful if the pattern only contains a few huge peaks.

Full Changelog: v1.1.3...v1.1.4

v1.1.3

What's Changed

• Load rasx files into Dara by @CompRhys in #24

New Contributors

• @CompRhys made their first contribution in #24

Full Changelog: v1.1.2...v1.1.3

v1.1.2

Update the tol in the export structure function, as BGMN only has 4 decimal places for atomic positions.

v1.1.1

Fix the bug of working_dir in ray. Previously, the job can raise FileNotFound when calling "os.getcwd" if ray was launched in a tmp folder (dara server runs all the jobs in the tmp folder)

v1.1.0

Introducing express mode

In this version, we introduce express mode to the search_phases routine. The option is designed to significantly reduce the computing time in the XRD phase analysis process, roughly 70~80%. This option is turned on by default.

Previous behavior

Previously, the grouping happened during the tree expansion. If multiple nodes happen to have a similar diffraction pattern and are sibling nodes, they will be merged together and only one will be expanded. The phase to be expanded is selected by a FoM to measure the fitness of the phases. The others will be listed as "alternative" phases to the expanded phase. This still requires many more refinements for every possible phase combination.

Current behavior

The way it works is by grouping the phases by diffraction similarity at the beginning of the search process. Only the phase with the top FoM in each group will be considered during the search. In this way, there will be much less phases used in the tree expansion process, thus, reducing the computing time to construct a search tree.

At the end of the search, the grouped phases will be added back as "alternative phases". We expect in most cases, results between express_mode=True and express_mode will be the same.

Usage

The express mode is enabled by default. To explictly enable it,

from dara import search_phases

result = search_phases(
    ...,
    express_mode=True
)

There will be no difference in other parameters and outputs.