MateriApps
A Portal Site of Materials Science Simulation

now 307 Apps

Inquiry / Application Request
Detailed search

XenonPy

  • Openness:3 ★★★
  • Document quality:2 ★★☆

XenonPy is a high-throughput material exploration framework based on machine learning technologies. This library can generate various chem/phys descriptors for machine learning to explore materials in virtual environment. Descriptors in matminer can be also used. Model training is done by PyTorch. Visualization tool for descriptor and transfer learning framework are also provided.

Long and Winding Road to XenonPy
Last Update:2021/12/09
You already voted!

Author: MateriApps Development Team (2020/2/19)

Introduction:

Xenon is a Python tool for searching for materials using machine learning. Here, we will install it on MateriApps LIVE!, and try the tutorials

Version of software

  • MateriApps LIVE! version 2.4 (Debian Buster, Python 3.5)
  • XenonPy version 0.3.1

Installation of XenonPy on MateriApps LIVE!:

VirtualBox settings

By default, MateriApps LIVE! is allocated 1GB of memory. As 1GB is not enough to install XenonPy, first we increase the memory allocation to 2GB

  • From the Oracle VM VirtualBox manager screen
  • Choose MateriAppsLive-2.4-amd64
  • “Setting” button → “System” button → “Motherboard” tag
  • Increase “Base memory” to 2 GB, and push “OK”

Below, we log in to MateriApps LIVE! and work in LX Terminal

Preparation of directories etc.

$ sudo mkdir -p /usr/local/lib/python3.5
$ sudo ln -s python3.5 /usr/local/lib/python3
$ echo '/usr/local/lib' | sudo tee /etc/ld.so.conf.d/local.conf
$ sudo ldconfig

RDkit for Python3

$ sudo apt -y install libboost-system-dev libboost-iostreams-dev libboost-python-dev libboost-serialization-dev
$ cd $ HOME
$ wget https://github.com/rdkit/rdkit/archive/Release_2019_09_3.tar.gz
$ tar zxvf Release_2019_09_3.tar.gz
$ cd rdkit-Release_2019_09_3
$ mkdir build
$ cd build
$ cmake -DPYTHON_EXECUTABLE = / usr / bin / python3 -DRDK_INSTALL_INTREE = OFF -DRDK_INSTALL_STATIC_LIBS = OFF ..
$ make
$ sudo make install

Upgrading Numpy

 $ sudo pip3 install -U numpy == 1.13.3

Cython

 $ sudo apt install cython

XenonPy

 $ sudo pip3 install --no-cache-dir xenonpy

Runing XenonPy tutorial:

Let’s run tutorials based on the XenonPy tutorial https://xenonpy.readthedocs.io/en/latest/tutorial.html

Getting Materials API key

  • Go to https://materialsproject.org
  • Click Login at the top right
  • Log in by Google account, Github account, or create an account with an email address and password
  • After logging in, click on the Dashboard in the upper right
  • Write down the key (character string of 16 characters) displayed under “Generate API Key”

Getting sample files

$ mkdir $HOME/XenonPy
$ cd $HOME/XenonPy
$ wget -O - https://github.com/yoshida-lab/XenonPy/archive/master.tar.gz | tar zxf - --strip-components=2 XenonPy-master/samples

Running the tutorials

$ cd $HOME/XenonPy
$ jupyter-notebook

Following the XenonPy tutorial https://xenonpy.readthedocs.io/en/latest/tutorial.html, please run the tutorials

TOP