It seems like many data science repos store their import package in the root of their distribution package, while it's more like a pypi package to put the import package into a src/ directory (there appear to be some benefits from a packaging/testing/islolation perspective?). Is it worth switching?
It seems like many data science repos store their import package in the root of their distribution package, while it's more like a pypi package to put the import package into a src/ directory (there appear to be some benefits from a packaging/testing/islolation perspective?). Is it worth switching?