☑️
FAIR test
fair-test is a library to build and deploy FAIR metrics tests APIs supporting the specifications used by the FAIRMetrics working group.
It aims to enable python developers to easily write, and deploy FAIR metric tests functions that can be queried by various FAIR evaluations services, such as FAIR enough and the FAIRsharing FAIR Evaluator
Feel free to create an issue, or send a pull request if you are facing issues or would like to see a feature implemented.
ℹ️
How it works
The user defines and registers custom FAIR metrics tests in separated files in a specific folder (the metrics folder by default), and start the API.
Built with FastAPI, pydantic and RDFLib. Tested for Python 3.7, 3.8 and 3.9
📥
Install the package
Install the package from PyPI:
pip install fair-test
🐍
Build a FAIR metrics test API
Checkout the example folder for a complete working app example to get started, including a docker deployment.
If you want to start from a project with everything ready to deploy in production we recommend you to fork the fair-enough-metrics repository.
📝
Define the API
Create a main.py file to declare the API, you can provide a different folder than metrics here, the folder path is relative to where you start the API (the root of the repository):
from fair_test import FairTestAPI
app = FairTestAPI(
title='FAIR Metrics tests API',
metrics_folder_path='metrics',
description="""FAIR Metrics tests API""",
cors_enabled=True,
license_info = {
"name": "MIT license",
"url": "https://opensource.org/licenses/MIT"
},
)
Create a .env file to provide informations used for the API, such as contact details and the host URL (note that you don't need to change it for localhost in development), e.g.:
HOST="metrics.api.fair-enough.semanticscience.org"
BASE_URI="https://metrics.api.fair-enough.semanticscience.org"
CONTACT_URL="https://github.com/MaastrichtU-IDS/fair-enough-metrics"
CONTACT_NAME="Vincent Emonet"
CONTACT_EMAIL="[email protected]"
CONTACT_ORCID="0000-0000-0000-0000"
ORG_NAME="Institute of Data Science at Maastricht University"
DEFAULT_SUBJECT="https://doi.org/10.1594/PANGAEA.908011"
🎯
Define a FAIR metrics test
Create a a1_my_test.py file in the metrics folder with your test:
from fair_test import FairTest
class MetricTest(FairTest):
metric_path = 'a1-check-something'
applies_to_principle = 'A1'
title = 'Check something'
description = """Test something"""
author = 'https://orcid.org/0000-0000-0000-0000'
metric_version = '0.1.0'
def evaluate(self):
self.info(f'Checking something for {self.subject}')
g = self.getRDF(self.subject, use_harvester=False)
if len(g) > 0:
self.success(f'{len(g)} triples found, test sucessful')
else:
self.failure('No triples found, test failed')
return self.response()
A few common operations are available on the
selfobject, such as logging or retrieving RDF metadata from a URL.
🦄
Deploy the API
You can then run the metrics tests API on http://localhost:8000 with uvicorn, e.g. with the code provided in the example folder:
cd example
pip install -r requirements.txt
uvicorn main:app --reload
Checkout in the
example/README.mdfor more details, such as deploying it with docker.
🧑💻
Development
📥
Install for development
Clone the repository and install the dependencies locally for development:
git clone https://github.com/MaastrichtU-IDS/fair-test
cd fair-test
pip install -e .
You can try to use a virtual environment to avoid conflicts, if you face issues:
# Create the virtual environment folder in your workspace
python3 -m venv .venv
# Activate it using a script in the created folder
source .venv/bin/activate
✔️
Run the tests
Install pytest for testing:
pip install pytest
Run the tests locally (from the root folder) and display prints:
pytest -s
📂
Projects using fair-test
Here are some projects using fair-test to deploy FAIR test services:
- https://github.com/MaastrichtU-IDS/fair-enough-metrics
- A generic FAIR metrics tests service developed at the Institute of Data Science at Maastricht University.
- https://github.com/LUMC-BioSemantics/RD-FAIRmetric-F4
- A FAIR metrics tests service for Rare Disease research.
4 Oct 20, 2022
4 Mar 26, 2022
23 Dec 1, 2022
49 Dec 23, 2022
1 Nov 12, 2021
6.2k Jan 8, 2023
756 Jan 6, 2023
1 Dec 20, 2021
1 Dec 22, 2021
4 Oct 11, 2021
30 Nov 08, 2022
4 Sep 02, 2022
3 Sep 30, 2021
3 Dec 20, 2022
3 Jul 28, 2022
2 Jun 30, 2022
1 Nov 28, 2021
13.5k Dec 01, 2021
3.4k Jan 01, 2023
109 Jan 07, 2023
3 Mar 21, 2022
38 Dec 25, 2022
657 Jan 02, 2023
2 Oct 11, 2022
9 Apr 15, 2022
25 Dec 23, 2022
39 Dec 31, 2022