Add environment.yml files per directory
#43
Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
|
| # This is a hack to install pyarrow via pip because libgdal-arrow-parquet constrains libarrow to | ||
| # v10, which constrains pyarrow to v10, which is too old for the `filesystem` parameter at the end | ||
| # of the file | ||
| - pip: | ||
| - pyarrow |
There was a problem hiding this comment.
This hack should be able to be removed after conda-forge/gdal-feedstock#796 is merged
|
Thanks @kylebarron I don't think I'm going to go through and test myself today but so far this looks good. It looks like this branch is behind main though, I was looking to add instructions about having conda or mamba installed in https://github.com/developmentseed/cloud-optimized-geospatial-formats-guide/blob/main/index.qmd#L14-L18 or contributing.md |
|
I added a sentence in the contributing docs that we expect an |
Change list
environment.ymlfile per directory containing the conda packages required for that notebook to run. To me this seems more maintainable than trying to include every dependency in one or two environment files, and seems more straightforward for users: "use the linked environment for each notebook"I went through and ran each notebook in each directory, verifying that they ran for me with these sets of dependencies.
Closes https://github.com/developmentseed/cloud-optimized-geospatial-formats-guide/issues/20