Closed NormannR closed 3 years ago
It is possible to run ipykernel within the poetry environment and the dependencies are properly shared. In VSCode, for example, choosing the poetry environment interpreter works well on my computer.
Ok, I'll test. I thought the ipykernels were shared by the session and indépendent from the environments (each kernel activating it's own environment)
On Mon, Dec 21, 2020, 5:44 PM Normann Rion notifications@github.com wrote:
It is possible to run ipykernel within the poetry environment and the dependencies are properly shared. In VSCode, for example, choosing the poetry environment interpreter works well on my computer.
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/econ-ark/DARKolo/pull/22#issuecomment-749070232, or unsubscribe https://github.com/notifications/unsubscribe-auth/AACDSKJDESSK22CD75ZSLO3SV53NPANCNFSM4VENQ5YQ .
I thought too, but it ended up working without manipulation. Otherwise, adding the environment to the kernelspec through a python -m ipykernel install --user --name=myenv
or modifying the kernel.json
file should do the trick.
I only modified the notebooks. If they are validated, I shall export them to the .py format before the PR is merged.
The pyproject.toml is a nice touch. It is a way to require the right version from dolo and dolark. But in this instance, it complicates stuff a little bit, because we have jupyter notebooks. The kernels need to have the right versions of the library, and I"m not sure how we can ensure that.