Raw File
Tip revision: 351b2193bdc2bd05fc295a85971436a8abba2c68 authored by Jakob Kraus on 07 October 2022, 12:12:09 UTC
Tip revision: 351b219
![GitHub Logo](/images/pyflosic_logo.png)

**Python-based Fermi-Löwdin orbital self-interaction correction (FLO-SIC)**   


The following document will guide you through the setup and show you how to get started with this Python-based FLO-SIC code.

#### Contents

- **: Installation guide in Markdown (md) language. 
- **: The authors of PyFLOSIC along with contact information.
- **: The version of PyFLOSIC you have acquired.
- *src/*: Contains the source files for PyFLOSIC.
- *examples/*: Contains useful examples to get started with PyFLOSIC.
- *doc/*: Can provide additional information on the code features. 
- *utils/*: Utilities that are useful for PyFLOSIC.
- *test/*: Contains a version control test.
- *LICENSE*: APACHE2 license file.

Please see the application examples in the folder examples for an in-depth introduction to the PyFLOSIC code.

The self-interaction error (SIE) is one of the mayor drawbacks of one of the most widley used electronic structure methods - density functional theory (DFT). Pederson et al. proposed a unitarily invariant and numerically feasible method based on Fermi-Löwdin orbital self-interaction correction (FLO-SIC). We implemented this method using the modern PySCF electronic structure code as basis.   

#### Theoretical foundations  
* Mark R. Pederson, Adrienn Ruzsinszky, and John P. Perdew. Communication: Self-interaction correction with unitary invariance in density functional theory. The Journal of Chemical Physics, 140(12):121103, March 2014.
* Zeng-hui Yang, Mark R. Pederson, and John P. Perdew. Full self-consistency in the Fermi-orbital self-interaction correction. Physical Review A, 95(5):052505, May 2017.  

## Installation 
You need a working pyscf installation on your system. 

## Dependencies (required)
We recommend the following package versions: 

* ase 3.17.0
* h5py 2.10.0
* numba 0.48.0
* scipy 1.5.2
* numpy 1.19.1
* pyscf 1.7.1

In order to make sure you use the recommended package versions for your PyFLOSIC calculations, you can type the following shell command:

$ source
This will create a virtual environment and install all of the required packages in the versions listed above.  
Most importantly, your previous installations of ase, h5py, etc. outside the virtual environment are not affected at all by this procedure.
If you want to leave the virtual environment, type the following shell command:

$ deactivate

## Dependencies (optional)

* matplotlib 
* python3-tk
* pyberny


## Authors 

* Sebastian Schwalbe ( 
* Kai Trepte (
* Susi Lehtola (
* Jens Kortus (

### previously 
* Jakob Kraus ( 
* Lenz Fiedler (
* Torsten Hahn (

The development of PyFLOSIC started with the master's thesis of Lenz Fiedler. Over the last year, we had many updates and complete code re-writes by Sebastian Schwalbe (ase backends, classes etc.) and Torsten Hahn (various speed-up techniques). Our main testers are Kai Trepte, Sebastian Schwalbe and Jakob Kraus. Jens Kortus is our overall theoretical guide and head of decisions, whereas Susi Lehtola has recently joined the team and provides new perspectives.

## Referencing this code
If you use the PyFLOSIC code for a scientific article or contribution, please cite the following article: 

* **PyFLOSIC: Python-based Fermi-Löwdin orbital self-interaction correction**  
  Sebastian Schwalbe, Lenz Fiedler, Jakob Kraus, Jens Kortus, Kai Trepte, and Susi Lehtola
  Journal of Chemical Physics 153:084104, August 2020. [DOI:10.1063/5.0012519]( [arXiv](
Applications of the PyFLOSIC code:     
* [Interpretation and Automatic Generation of Fermi‐Orbital Descriptors](
* [Stretched or noded orbital densities and self-interaction correction in density functional theory](
## Milestones 
  * **April 30, 2020**   
Thanks to all the people who have read our article. We are happy that our article "Interpretation and Automatic Generation of Fermi‐Orbital Descriptors", where every results was produced with our PyFLOSIC code, published in the Journal of Computational Chemistry (JCC) was promoted to one of the top 10% downloaded articles. [1]( 
  * **August 24, 2020**  
We are very happy to announce that our PyFLOSIC article (see above) is now finally published in the Journal of Chemical Physics (JCP)! 
  * **September 01, 2021**     
We are pleased that our PyFLOSIC JCP article was highlighted by JCP. 

## Some Remarks
PyFLOSIC is still a relatively young code subject to regular significant changes.  If you discover any issues while working with PyFLOSIC, feel free to contact us. 
back to top