Using Embedded Python with virtual environments and with Jupyter can be a little tricky because, in part, you need to start python by running the irispython command instead of python3.
This repository is an example of how to create a container that enables python-first development with embedded python, creates a virtual environment named embedded python and creates an ipython kernel (which are used in jupyter) that also uses embedded python.
In this directory, run the following:
docker build --tag iris-embedded-jupyter .
Chances are, you're running this container because you want to run Jupyter with embedded python. That means you're going to want to run both the IRIS server on port 52773 and jupyter on port 8888.
The most basic way to run the container is:
docker run -p 9999:52773 -p 8888:8888 iris-embedded-jupyter --check-caps false
docker exec -it <containerID> ./start-jupyter.sh