Metadata-Version: 2.1
Name: nomad-dos-fingerprints
Version: 1.0.2
Summary: An implementation of DOS fingerprints for the NOMAD Laboratory.
Home-page: https://gitlab.mpcdf.mpg.de/nomad-lab/nomad-dos-fingerprints
Author: The NOMAD Authors
Author-email: kuban@physik.hu-berlin.de
License: Apache-2.0
Project-URL: homepage, https://github.com/nomad-coe/nomad-dos-fingerprints
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: nomad-lab >=1.3.6.dev1
Provides-Extra: dev
Requires-Dist: mypy ==1.0.1 ; extra == 'dev'
Requires-Dist: pytest ==3.10.0 ; extra == 'dev'
Requires-Dist: pytest-timeout ==1.4.2 ; extra == 'dev'
Requires-Dist: pytest-cov ==2.7.1 ; extra == 'dev'
Requires-Dist: ruff ==0.1.4 ; extra == 'dev'
Requires-Dist: uv >=0.2.35 ; extra == 'dev'

This package implements fingerprints of the electronic density-of-states (DOS) for the evaluation of similarity of materials based on their electronic structures.

The fingerprints are based on a modification on the D-Fingerprints presented in Ref. [1].
Our modification allows to target specific energy ranges for the evaluation of the similarity of the electronic structure.
As a similarity measure we use the Tanimoto coefficient [2].

# Usage

Fingerprints are instances of the `DOSFingerprint()` class and can be calculated by providing the energy in [Joule] and the DOS in [states/unit cell/Joule] to the `calculate()` method. Furthermore, the parameters of a non-uniform grid can be chosen. The default grid is specialized on the energy range between -10 and 5 eV and emphasizes the upper valence region.

```Python
from nomad_dos_fingerprints import DOSFingerprint
dos_fingerprint = DOSFingerprint().calculate(<dos_energies>,<dos_values>)
```

To evaluate the similarity, the function `tanimoto_similarity()` can be used:

```Python
from nomad_dos_fingerprints import tanimoto_similarity
tc = tanimoto_similarity(dos_fingerprint_1, dos_fingerprint_2)
```

# References

[1] Isayev _et al._, Chem. Mater. 2015, 27, 3, 735–743 (doi:10.1021/cm503507h)

[2] P. Willet _et al._, J. Chem. Inf. Comput . 38 , 983 996 (1998) (doi:10.1021/ci9800211)
