Pyntfa

Description

*Pyntfa is a python package for non-stationary time-frequency analysis. This package has a variety of applications in both exploration and earthquake seismology. The initial version of this package was held at https://github.com/chenyk1990/pyntfa, which is no longer maintained since 11/25/2024.

Reference

Chen, Y., 2021, Nonstationary local time-frequency transform, Geophysics, 86(3), V245–V254.

Chen, Y., et al., 2025, Pyntfa: A python package for non-stationary time-frequency analysis, TBD.

BibTeX:

@article{ntfa,
  title={Nonstationary local time-frequency transform},
  author={Yangkang Chen},
  journal={Geophysics},
  volume={86},
  number={3},
  issue={3},
  pages={V245–V254},
  doi={10.1190/geo2020-0298.1},
  year={2021}
}

@article{pyntfa,
  title={PYntfa: A python package for non-stationary time-frequency analysis},
  author={Yangkang Chen and co-authors},
  journal={TBD},
  volume={TBD},
  number={TBD},
  issue={TBD},
  pages={TBD},
  doi={TBD},
  year={2025}
}

---

Copyright

The pyntfa developing team, 2022-present

---

License

GNU General Public License, Version 3
(http://www.gnu.org/copyleft/gpl.html)   

---

Install

Using the latest version (Suggested)

git clone https://github.com/aaspip/pyntfa
cd pyntfa
pip install -v -e .

or using Pypi

pip install pyntfa

or

pip install git+https://github.com/aaspip/pyntfa

---

Examples

The "demo" directory contains all runable scripts to demonstrate different applications of pyntfa. 

---

Dependence Packages

• scipy
• numpy
• matplotlib
• pynpre

---

Modules

ntfa.py  -> core python fuction for calling the Cpython (ntfacfun)

---

Development

The development team welcomes voluntary contributions from any open-source enthusiast. 
If you want to make contribution to this project, feel free to contact the development team. 

---

Contact

Regarding any questions, bugs, developments, collaborations, please contact  
Yangkang Chen
chenyk2016@gmail.com

---

Gallery

The gallery figures of the pyntfapackage can be found at https://github.com/aaspip/gallery/tree/main/pyntfa Each figure in the gallery directory corresponds to a DEMO script in the "demo" directory with the exactly the same file name.

The following figure shows a 1D synthetic example using the pyntfa package. Generated by demos/test_pyntfa_chirp.py

We can further boost the resolution by the following example by

demos/test_pyntfa_chirp.py

A comparion between CWT and NTFA via a noise test (this example also reproduces the Figure 4 in Zhang et al., 2020, Convolutional neural networks for microseismic waveform classification and arrival picking, Geophysics, 85, WA227-WA240). Generated by

demos/test_pyntfa_noise_cwtVSntfa.py

The following figure shows another example on strongly non-stationary 1D synthetic signal using the pyntfa package. Generated by demos/test_pyntfa_chao.py

The following figure shows an application on one USArray trace (This example reproduces the Figure 5&6 in Chen, 2021, Nonstationary local time-frequency transform, Geophysics, 86(3), V245–V254.) demos/test_pyntfa_usarray1d.py

The following figure shows an application on the LIGO data (for detecting gravitational wave signal) demos/test_pyntfa_ligo.py

The following figure shows an application on the PcP phase (for studying core-mantle boundary) demos/test_pyntfa_pcp.py

The following figure shows an application on the PcP phase with STFT and ST. The comparisons with STFT/ST can be done by uncommenting the corresponding code of demos/test_pyntfa_pcp.py A general observation is that while STFT fails to extract the PcP phase distinctly from other phases and noise, ST is inferior to NTFA in that it is less focused on primary phases (P and PcP) and has weaker anti-noise ability.

The following figure shows an application on the SS precursor (for studying MTZ) demos/test_pyntfa_ss.py