Extract Grid class from jellium_model's signature and use it in a few methods

[Strilanc]

I'll use it in far more places after this gets merged.

[Strilanc]

@babbush Changing all those things at once will create a gigantic change that's too hard to review. We have to bite a piece at a time.

Point at the examples that don't work and I'll fix those, but we should defer consistency to future PRs making more and more things use Grid.

[babbush]

Now that I look I think none of the examples in the docs break actually. But there is a section called "Hamiltonians in the plane wave basis" in the ipython notebook in examples/. Update that and I'll approve the PR. Unless you are going to update the Hubbard code right away it would be helpful if you could also open an issue that says "use Grid class in Hubbard code".

[babbush]

Also, I just want to draw this PR to the attention of @Spaceenter who has been helping me with the code related to this plane wave stuff.

[Strilanc]

How am I supposed to update the notebook? We're including hard-coded output in the repo?

I updated the example code, but whenever I rerun anything it just complains about fermilib not being found.

[babbush]

Yes, the reason we have the output saved in the notebook is so that people can see the input code and its output on GitHub without needing to actually open the file. Are you saying that the ipython notebook is not running on your machine? That sounds like an issue...

[Strilanc]

Well, I've never used a jupyter notebook before so it's not too surprising that there's issues. For example, I probably need to add a path argument or tell it the virtualenv to use or something like that.

[babbush]

Ah ok. They are all the rage in science these days. Especially since I screwed us all by accidentally accepting that non-CLA pull request and now need to wait for Damian to revert, maybe try to get the jupyter notebooks working.

[damiansteiger]

The non-CLA pull request has been reverted. New merges are again possible. Please note that the develop branch is now the default branch (i.e., pull request should still be opened to the develop branch)

[Strilanc]

The commit is still in the history. Is it sufficient to revert it?

[damiansteiger]

We will check if a hard reset is possible on GitHub

[Strilanc]

It's possible, but you'll have to disable the protected status of master while you do it. So:

• Fix the issue on your local copy.
• Turn off master protection.
• Force push master
• Turn on master protection

[Strilanc]

Squash-rebased.

[damiansteiger]

Thanks pointing this out. We have performed a hard reset.

I am very sorry that you had to rebase your pull request in order to remove this commit as well.