Metadata-Version: 2.1
Name: ESPNN
Version: 1.0.1
Summary: Electronic Stopping Power Neural Network predictor
Home-page: https://github.com/ale-mendez/SPNN
Author: 
    Felipe Bivort Haiek,
    Alejandra Mendez,
    Claudia Montanari,
    Darío Mitnik
    
Author-email: felipebihaiek@gmail.com, alemdz.7@gmail.com
License: The GPLv3 License
Keywords: Stopping Power,Energy Loss,Atoms,Molecules,Ions,Machine Learning,Neural Network
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Education
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: GNU General Public License v3 (GPLv3)
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3.7
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Topic :: Scientific/Engineering
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: matplotlib==3.5.2
Requires-Dist: numpy==1.21.6
Requires-Dist: pandas==1.3.5
Requires-Dist: pyvalem==2.5.7
Requires-Dist: torch==1.11.0
Requires-Dist: scikit-learn==0.24.2
Requires-Dist: joblib==1.0.0

# ESPNN - Electronic Stopping Power Neural Network

 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT) [![develstat](https://github.com/ale-mendez/ESPNN/actions/workflows/espnn_ci.yml/badge.svg)](https://github.com/ale-mendez/ESPNN/actions/workflows/espnn_ci.yml/badge.svg) [![codecov](https://codecov.io/gh/ale-mendez/ESPNN/branch/master/graph/badge.svg?token=R49KN0O0I1)](https://codecov.io/gh/ale-mendez/ESPNN) [![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/ale-mendez/ESPNN/master?urlpath=%2Fdoc%2Ftree%2F%2Fworkflow%2Fprediction.ipynb)
 <!-- [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1UCDj0XT_4Ex_Mvp1vurleeeDVcjed6vP) -->
 <!-- [![Research software impact](http://depsy.org/api/package/pypi/)](http://depsy.org/package/python/) -->

The ESPNN is a Python-based deep neural network that enables users to predict the electronic stopping power cross-section for any ion and target[^1] combinations for a wide range of incident energies. The deep neural network was trained on tens of thousands of curated data points from the [IAEA database](https://www-nds.iaea.org/stopping/). See more details of the ESPNN in this [publication](https://github.com/ale-mendez/ESPNN-doc).

 <!--
### Citation

```
@article{BivortHaiek2022,
author = {F. Bivort Haiek, A. M. P. Mendez, C. C. Montanari, D. M. Mitnik},
title = {ESPNN: The IAEA stopping power database neutral network. Part I: Monoatomic targets.},
year = {2022}

```
}-->

You can use the ESPNN package [remotely](#run-ESPNN-online) or [locally](#install-espnn). Find below all the usage options available.

If you encounter problems installing the package or notice troubling features in the stopping power model, make sure to post an [issue](https://github.com/ale-mendez/ESPNN/issues) or send us an email[^2].

## Run ESPNN online

The ESPNN package can be used remotely in the <a href="https://mybinder.org/v2/gh/ale-mendez/ESPNN/master?urlpath=%2Fdoc%2Ftree%2F%2Fworkflow%2Fprediction.ipynb" target="_blank">Binder</a> platform. There, you'll find a Jupyter notebook with a quick tutorial on how to use the ESPNN. You can compute the stopping power of any projectile-target combination in that Jupyter notebook. The stopping power results are saved in plain text files and can be downloaded by clicking on the folder icon in the vertical left menu. *Make sure to download them as they will be lost once the tab is closed.*

## Install ESPNN

To use the ESPNN on your computer, first, you'll need to install it. ESPNN is currently restricted to Python 3.7 and 3.8. We recommend using a Python virtual environment to this end (for example, see <a href="https://docs.anaconda.com/anaconda/install/index.html" target="_blank">anaconda</a> or <a href="https://virtualenv.pypa.io/en/stable/installation.html" target="_blank">virtualenv</a>). If you are not familiar with virtual environments and would like to rapidly start using Python, follow the <a href="https://docs.anaconda.com/anaconda/install/index.html" target="_blank">Anaconda</a> instructions according to your operating system:

- <a href="https://docs.anaconda.com/anaconda/install/linux/" target="_blank">Install anaconda in Linux</a>
- <a href="https://docs.anaconda.com/anaconda/install/windows/" target="_blank">Install anaconda in Windows</a>
- <a href="https://docs.anaconda.com/anaconda/install/mac-os/" target="_blank">Install anaconda in macOS</a>

### Using pip

The simplest way to install the ESPNN is via pip. Indistinctly, Ubuntu, Windows, and macOS users can install the package by typing in the terminal or the Anaconda bash terminal:

```console
pip install ESPNN
```

### Using this repository

You can also install the ESPNN package by cloning or [downloading](https://github.com/ale-mendez/ESPNN/archive/refs/heads/master.zip) this repository. To clone (make sure you have git installed) this repo, use the following commands in your terminal/anaconda bash terminal:

```console
git clone https://github.com/ale-mendez/ESPNN.git
cd ESPNN
pip install ESPNN/
```

If you [downloaded](https://github.com/ale-mendez/ESPNN/archive/refs/heads/master.zip) the zip, change your directory to your download folder and, in your terminal/anaconda bash terminal, type

```console
pip install ESPNN-master.zip
```

## Run ESPNN locally

Once you've [installed](#install-espnn) the ESPNN package in your preferred environment, you can run it by using a jupyter notebook or directly from terminal.

### Using a notebook

A basic tutorial of the ESPNN package usage is given in <a href="https://github.com/ale-mendez/ESPNN/blob/master/workflow/prediction.ipynb" target="_blank">prediction.ipynb</a>. The package requires the following parameters:

- ``projectile``: Chemical formula for the projectile
- ``target``: Chemical formula for the target

```python
import ESPNN
ESPNN.run_NN(projectile='He', target='Au')
```

![](https://github.com/ale-mendez/ESPNN/blob/master/docs/prediction_files/prediction_2_0.png?raw=true)

The package automatically produces a ``matplotlib`` figure and a sample file named ``XY_prediction.dat``, where ``X`` is the name of the projectile and ``Y`` is the name of the target system.

```console
ls -a
.  ..  HHe_prediction.dat  prediction.ipynb 
```

#### Optional arguments

The energy grid used for the ESPNN calculation can be customized with arguments

- ``emin``: Minimum energy value in MeV/amu units (default: ``0.001``)
- ``emax``: Maximum energy value in MeV/amu units (default: ``10``)
- ``npoints``: Number of grid points (default: ``150``)

Furthermore, the figure plotting and output-file directory path can be modified via

- ``plot``: Prediction plot (default: ``True``)
- ``outdir``: Path to output folder (default: ``"./"``)

```python
ESPNN.run_NN(projectile='H', target='Ta', emin=0.0001, emax=100, npoints=200)
```

![](https://github.com/ale-mendez/ESPNN/blob/master/docs/prediction_files/prediction_4_0.png?raw=true)

### From terminal

The ESPNN package can also be used from the terminal with a syntax analogous to the above given:

```console
python -m ESPNN H Au
```

Additional information about the optional arguments input can be obtained with the -h, --help flag:

```console
python -m ESPNN -h
```

## Funding Acknowledgements

The following institutions financially support this work: the Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) by the PIP-11220200102421CO and the Agencia Nacional de Promoción Científica y Tecnológica (ANPCyT) of Argentina PICT-2020-SERIEA-01931. CCM also acknowledges the financial support of the IAEA.

[^1]: *ESPNN first release considers only mono-atomic targets.*
[^2]: felipebihaiek@gmail.com, alemdz.7@gmail.com
