A simple example with numpy dependencies and using pycape and serdio.lift_io to handle automatic serialization/deserialization of Cape function input/outputs. All commands are run from the repo root directory.
Note the numpy_serde.py helper, which defines a custom encoder/decoder bundle that allows serdio to handle numpy arrays.
Build the deployment package:
# Create a deployment folder
mkdir np-stats-deployment
# Copy the cape function (app.py) in the deployment folder
cp app.py numpy_serde.py np-stats-deployment/.
# Add serdio and numpy dependencies using docker
docker run -v `pwd`:/build -w /build --rm -it python:3.9-slim-bullseye pip install -r requirements.txt --target np-stats-deployment/Deploy the function:
Deploy with the CLI the function as follow:
cape deploy np-stats-deployment
export FUNCTION_ID=<function id from above>
Generate the function token based on the function ID and function checksum returned by deploy:
cape token create --name np-stats
export TOKEN=<token from above>
Run the function: You can run the function with PyCape as follow:
python run.py