Skip to content

Latest commit

 

History

History
executable file
·
39 lines (33 loc) · 2.17 KB

paper.md

File metadata and controls

executable file
·
39 lines (33 loc) · 2.17 KB
title tags authors affiliations date bibliography
MPDS platform Python API client
materials-science
materials-informatics
materials-platform
crystallography
crystal-structure
phase-diagram
data-science
Python
name orcid affiliation
Evgeny Blokhin
0000-0002-5333-3947
1, 2
name affiliation
Martin Uhrin
3
name index
Tilde Materials Informatics, Berlin, Germany
1
name index
Materials Platform for Data Science, Ltd., Tallinn, Estonia
2
name index
Ecole polytechnique federale de Lausanne, Lausanne, Switzerland
3
19 July 2018
paper.bib

Summary

The MPDS platform [@mpds] is an online database for inorganic chemistry and materials science with nearly two million entries: physical properties, crystalline structures, phase diagrams etc., available via API, ready for the modern data-intensive applications. The source of the data are nearly 300 thousands of peer-reviewed articles, books and conference proceedings, published since 1891 until today.

To work with the MPDS platform API, any programming language, able to execute HTTP requests and handle the JSON output, can be employed. However, one of the most frequently used languages for data processing is Python. Therefore we provide a client library for Python (versions 2.7 and 3.6). Installed as any other Python package via pip, this client library takes care of many data retrieval aspects, such as querying, pagination, error handling, response validation, proper data extraction and more. Additionally, even more advanced transformations can be done with the aid of this library [@labs], such as calculating the crystalline descriptors for machine learning, sample ordering of the disordered alloy structures etc. The MPDS platform API client library was created as a part of the matminer package [@Ward:2018] for data mining in materials science (LBNL, Berkeley, USA) and then became a standalone package.

We encourage the MPDS platform users to adopt this library for building their own materials data infrastructures.

The source code for the MPDS platform Python API client has been archived to Zenodo with the linked DOI: [@zenodo]

References