Conversation
| run: | | ||
| conda install numpy cython pip hdf5 libnetcdf cftime netcdf4 --strict-channel-priority | ||
| pip install -e . --force-reinstall | ||
| conda env create -f environment.yml |
There was a problem hiding this comment.
Use the MHKiT conda env from enviroment.yml file
| conda activate mhkit-env | ||
| conda install -y pytest coverage coveralls |
There was a problem hiding this comment.
install development specific packages
| run: | | ||
| conda install numpy cython pip pytest hdf5 libnetcdf cftime netcdf4 coverage --strict-channel-priority | ||
| pip install -e . --force-reinstall | ||
| conda env create -f environment.yml |
There was a problem hiding this comment.
Use the MHKiT conda env from enviroment.yml file
| shell: bash -l {0} | ||
| run: | | ||
| conda activate mhkit-env | ||
| conda install -y pytest coverage coveralls nbval jupyter utm folium |
There was a problem hiding this comment.
Extra packages this time for the notebook examples
| @@ -1,4 +1,4 @@ | |||
| name: myenv | |||
| name: mhkit-env | |||
There was a problem hiding this comment.
set env name to mhkit-env
| - pecos>=0.3.0 | ||
| - notebook | ||
| - matplotlib>=3.9.1 | ||
| - fatpack | ||
| - nrel-rex |
There was a problem hiding this comment.
Solve everything using conda (no pip)
There was a problem hiding this comment.
the folium package updated and required a modification to the pacwave map
|
@akeeste this PR is ready for review |
There was a problem hiding this comment.
Would it be possible to install mhkit in this file at the end? It saves the installer a step? Not sure if you are not installing mhkit here on purpose?
There was a problem hiding this comment.
It would be possible and I am not installing it on purpose because it would not make sense for testing.
If MHKiT was included in the .env file then that would install the current conda release vs the changes in the repo we are trying to test.
My thoughts on how users would interact with the .env file:
- Anyone simply wanting to use MHKiT I expect to use our release via
conda install mhkit. - Anyone using the
.envfile I expect is doing development in the repo.
There was a problem hiding this comment.
That makes sense to me. I think our documentation could use an update on clearly delineating the use of the different install methods (environment.yml, conda install, editable) for different use cases (developer, user, user who wants notebooks too). I can take that on in the docs repo as I make other updates
- Anyone using the
.envfile I expect is doing development in the repo.
So the developer installation workflow would be:
- create conda environment with environment.yml
pip install -e .for a editable MHKiT install
Also pip install -e . is being deprecated and we'll need a new method soon
There was a problem hiding this comment.
That makes sense to me. I think our documentation could use an update on clearly delineating the use of the different install methods (environment.yml, conda install, editable) for different use cases (developer, user, user who wants notebooks too). I can take that on in the docs repo as I make other updates
- Anyone using the
.envfile I expect is doing development in the repo.
So the developer installation workflow would be:
- create conda environment with environment.yml
pip install -e .for a editable MHKiT install
Also pip install -e . is being deprecated and we'll need a new method soon
|
@akeeste I added the docstring back that I mistakenly deleted. Assuming the tests pass is this ready to merge? |
3 participants
Improve PyLint Compliance and Update GitHub Actions Environment