ReFANN (Reconstruct Functions with Artificial Neural Network)
ReFANN is a nonlinear interpolating tool based on ANN, without assuming a model or parameterization. It can reconstruct functions from data with no assumption to the data, and is a completely data-driven approach.
It is proposed by Guo-Jian Wang, Xiao-Jiao Ma, Si-Yao Li, Jun-Qing Xia (2020).
The documentation can be found at refann.readthedocs.io.
If you use this code in your research, please cite Guo-Jian Wang, Xiao-Jiao Ma, Si-Yao Li, Jun-Qing Xia, ApJS, 246, 13 (2020).
You can install refann by using pip:
$ sudo pip install refann
or from source:
$ git clone https://github.com/Guo-Jian-Wang/refann.git $ cd refann $ sudo python setup.py install
Copyright 2020-2020 Guojian Wang
refann is free software made available under the MIT License. For details see the LICENSE file.