Adapt for online lab compatibility

[juntyr]

With these changes, the following top-to-bottom example works: https://gist.github.com/juntyr/ea85adab5df19141a84939f2d322697f

[juntyr]

I'm definitely amenable to a different name, this just makes things work with the least effort without making anything worse

[treigerm]

Thanks @juntyr ! Looks good to me just had some very minor questions, feel free to merge when you want to.

Overall, the structure of having the scripts as modules of the libraries seems a bit unusual (in the sense that I don't think I have seen Python libaries do this before). But it should be fine for the EGU demo. Long term we should probably think about how we can refactor the scripts into some re-usable library components.

[juntyr]

Overall, the structure of having the scripts as modules of the libraries seems a bit unusual (in the sense that I don't think I have seen Python libaries do this before). But it should be fine for the EGU demo. Long term we should probably think about how we can refactor the scripts into some re-usable library components.

What if we call the module something other than scripts ...

On a more serious note, I feel like the compression and metric modules would not change much under a different structure (and we already use the executable module approach in the data loader). The error bound generation and summary printing are definitely more in script land.

For this PR, I just wanted to ensure that code importing the library can reuse all of the functionality. I agree that we can probably find a better way to organize them. If you can think of something, I can make the changes now, otherwise we can come back to it later

[treigerm]

No feel free to merge, I was mainly just thinking out loud. Feel free to merge!