Adding adcc

[maxscheurer]

Checklist

• Title of this PR is meaningful: e.g. "Adding my_nifty_package", not "updated meta.yaml".
• License file is packaged (see here for an example).
• Source is from official source.
• Package does not vendor other packages. (If a package uses the source of another package, they should be separate packages or the licenses of all packages need to be packaged).
• If static libraries are linked in, the license of the static library is packaged.
• Build number is 0.
• A tarball (url) rather than a repo (e.g. git_url) is used in your recipe (see here for more details).
• GitHub users listed in the maintainer section have posted a comment confirming they are willing to be listed there.
• When in trouble, please check our knowledge base documentation before pinging a team.

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-linting service.

I wanted to let you know that I linted all conda-recipes in your PR (recipes/adcc) and found some lint.

Here's what I've got...

For recipes/adcc:

• If python is a host requirement, it should be a run requirement.

For recipes/adcc:

• Recipes should usually depend on matplotlib-base as opposed to matplotlib so that runtime environments do not require large packages like qt.
• Whenever possible python packages should use pip. See https://conda-forge.org/docs/maintainer/adding_pkgs.html#use-pip
• License is not an SPDX identifier (or a custom LicenseRef) nor an SPDX license expression.

Documentation on acceptable licenses can be found here.

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipes/adcc) and found it was in an excellent condition.

[maxscheurer]

@conda-forge/staged-recipes ready for review 😄

[dopplershift]

Why run a test before building? That test could easily be done as part of the standard test phase.

[maxscheurer]

The tests are of course not run before building. Before building there would be no package to test.

The reason why it's not done in the usual test phase is that we don't install our package tests (they automatically download the reference data from our server, which we don't want to enable in every user's installation).

I'm thinking about a solution that everyone is happy with... Maybe we could patch in a minimal run test for adcc (like one of the examples), to make sure that it works beyond import. Other ideas, @mfherbst?

[mfherbst]

We could have a restricted set of tests, that just run a calculation for a small basis set and checks the excitation energies (plain values) agree and maybe that the excitation vectors have a small residual. Than we don't have to download testdata at all, but still check a large part of the code is functioning properly. I would not patch that in, rather fix that directly upstream and just call pytest appropriately here.

How does that sound to everyone?

[maxscheurer]

I think this is a good suggestion. So then we include this in upstream adcc, release v0.15.11 , and then continue here 😄

[maxscheurer]

@dopplershift We have adapted our test setup for c-f according to your suggestions (minimal tests can now be run during intended test phase). If you are happy with the current setup in general, we can release v0.15.11 such that all pipelines here will work.

[dopplershift]

The reason why it's not done in the usual test phase is that we don't install our package tests (they automatically download the reference data from our server, which we don't want to enable in every user's installation).

Just to clarify a point here. The standard test phase in the recipe is only run while running conda build, but after the package build. So it's not run on every users' system, only (generally) on conda-forge's CI systems.

In general, though, a reduced test set is great because you really don't need a full set here (that's what upstream's CI setup is for), just enough to validate that everything is built and linked correctly.

[maxscheurer]

So it's not run on every users' system, only (generally) on conda-forge's CI systems.

Yes, I know. I meant we don't want to install the entire test suite on users' machines, which would enable them to (accidentally) over and over download testdata from our server. Sorry for the misunderstanding 😉
If you are comfortable with the current setup, we can deploy the most recent version (which has minimal tests) and fix pipelines here.

[maxscheurer]

@dopplershift, pipelines are working fine now, including minimal tests. So if we could get another review, that'd be great 👍

[maxscheurer]

@conda-forge/staged-recipes @dopplershift ready for review 🚀

[maxscheurer]

pinging @conda-forge/staged-recipes @dopplershift again, would be cool if this could be re-reviewed/merged at some point 😄 Thanks! 👍