Skip to content

A code to estimate transport coefficients from the cepstral analysis of a (multi)variate current stationary time series -- [FKA "thermocepstrum"]

License

Notifications You must be signed in to change notification settings

sissaschool/sportran

Repository files navigation

SporTran (FKA thermocepstrum)

A code to estimate transport coefficients from the cepstral analysis of a multi-variate current stationary time series.

PyPI version Documentation Status

Documentation

https://sportran.readthedocs.io

References

Developed by Loris Ercole, Riccardo Bertossa, Sebastiano Bisacchi under the supervision of prof. Stefano Baroni

Acknowledgment The development of this software is part of the scientific program of the EU MaX Centre of Excellence for Supercomputing Applications (Grant No. 676598, 824143) and has been partly funded through it.


Usage

There is a GUI that you can try after installing the package. Click here for instructions.

The code can be used as a library, for example in a Jupyter notebook. In the examples folder you can find some examples.

Alternatively, you can run the code analysis.py from the command line without any installation procedure. It can execute most of the cepstral analysis routines, returning the results in a series of data files and PDF plots. See the examples/example_commandline_NaCl folder and the help (python analysis.py --help) for more information.

Requirements

  • numpy
  • scipy
  • matplotlib
  • tkinter
  • markdown2
  • pillow

Installation

You can simply pip-install SporTran downloading it from PyPI with pip install sportran.

Alternatively:

  1. Clone this repository: git clone https://github.com/sissaschool/sportran.git
  2. Install the package with pip (dependencies will be automatically downloaded). For example:
cd sportran
pip install .

You are all set! You can check that the installation is working by trying to run the command sportran-analysis.

The Graphical User Interface can be started with the command sportran-gui.

Issues

You are strongly encouraged to report any issue on the official GitHub issues page.