This is a toy data science project which illustrates how Docker and Makefiles can be used to achieve reproducibility. It's based on Cookiecutter Data Science.
From the current directory, run
docker build -t myproject .
to build the Docker image for the project. Then run
docker run -it myproject
to start a container.
From inside the container, you can do
make features to process the data
make train to train the model
make predict to obtain model predictions
make figures to produce the figures
make reports to produce the report