SeJV / MaterialsCombiner

Combines Apis of multiple materials databases

MaterialsCombiner

Combines Apis of multiple materials databases Currently included DataBases:

• Materials Project: from pymatgen import MPRester
• Mendeleev: from mendeleev import element
• OQMD: import qmpy_rester as qr

Start the service

• python>=3.7
• Install Requirements: pip install ./requirements.txt
• Add .env File on root directory, with the Key for Materials Project: MP_KEY=<your_key>
• Run with python3 main.py

Use the service

• on port:5000 the service is now available
• get information with a GET request on path /formula/<your_input>
  • Information gets cached locally, so the next time you reqest the same data, it will be much faster
  • Additionally you can use: '-' for OR, ',' for AND, e.g. (Fe-Mn),O for the OQMD database

Adding databases

• Additional databases with API
  • With Flask requests, any available API that has a formula as Input and returns an json object is addable
  • In the main.py inside the get function, there is an inp parameter, that represents the formula as string
  • This can be used to fetch a request, and results of such a request should be added to the data_in_json object
• Additional databases with python package
  • Similar to API approach, you can add databases that are available as python packages, such as
    • AFLOW: https://pypi.org/project/aflow/
    • Citrination: https://citrineinformatics.github.io/python-citrination-client/index.html
    • COD: https://pypi.org/project/crystals/
    • ICSD: https://github.com/hegdevinayi/icsd-queryer
  • Those are easier to implement, because you don't need an additional Flask app, but can use the already implemented functions
  • If the response format from the api is not available in json, you need to transform it to json
  • Concatenate the results into the data_in_json object in the main.py file inside the get function, with the key as the identifier for the database