search

ericmjl / bayesian-stats-modelling-tutorial

How to do Bayesian statistical modelling using numpy and PyMC3
MIT License
659 stars 280 forks source link

MNT: Add symbolic link to environment.yml for user clarity #104

Open matthewfeickert opened 2 years ago

matthewfeickert commented 2 years ago

This PR partially reverts things I did in PR #85. To make it easier for users who are not used to binder but maybe used to typical Conda workflows, add a symbolic link from binder/environment.yml to environment.yml in the top level of the repository. Additionally, change all instances of binder/environment.yml to environment.yml (that is, run

git grep --name-only "binder/environment.yml" | xargs sed -i 's|binder/environment.yml|environment.yml|g'

).

This does not fully revert the work of PR #85 though, as Binder will always look for config files under the binder/ directory in a repository if the directory exists. So this is more of a way to keep Binder config clear but also make things easier/clearer for users.

For an example of this behavior, c.f. https://github.com/gradhep/differentiable-analysis-examples/pull/1.

matthewfeickert commented 2 years ago

@ericmjl @hugobowne this is ready for review, though perhaps we should get the CI running in another PR and then come back and rebase this once that is done. Though as it seems the CI isn't supporting Windows I don't think we need to be worried about symbolic links failing there so this should be fine if you want to visually review only.

matthewfeickert commented 2 years ago

Gentle ping on this to see if there is anything else that needs to get done before merge.