Managing a ML pipeline with Luigi.
Case study: reports of the Italian COVID-19 situation.
We use the data of the Italian Civil Protection, available at this repository.
Clone the respository:
git clone https://github.com/elianap/luigi-covid-pipeline.git
You can firstly create an environment, for example using conda:
conda create --name luigi3 python=3
Then install the required package (in requirements.txt) via pip.
To install luigi:
pip install luigi
Run luigi scheduler
PYTHONPATH='.' luigid
Run Luigi task
luigi --module <module_name> <task_name> --<parameter1_name> <par_value>
Generate a daily report:
luigi --module covid_pipeline AggregateInReport