Adapt the plotting function to a standalone docker service. It should maintain the same input json of bias scores and save the rendered images to a shared volume using a standardized name. This is mainly a repackaging and should not change the existing code in huge ways.
Tasks
Create dockerfile under ./docker/plotter/
[x] Wrap the existing plotting code in a simple flask API that accepts POST requests with the existing input format.
[x] Change the image path to save to a local volume rather than S3
[x] Add the plotter dockerfile to the configuration in docker-compose.yaml
[ ] Write unit tests for the basic use case, these can be expanded later.
ivyleavedtoadflax
commented
5 years ago
Status
Assigning to @ivyleavedtoadflax Please use the https://github.com/N2ITN/are-you-fake-news/tree/develop-dockerize branch.
Issue
Adapt the plotting function to a standalone docker service. It should maintain the same input json of bias scores and save the rendered images to a shared volume using a standardized name. This is mainly a repackaging and should not change the existing code in huge ways.
Tasks
Create dockerfile under
./docker/plotter/