Fix issue #19: when input data has no non-co2 variables

[jkikstra]

For full description of error, see #19.

• Tests added
• Documentation added/changed if necessary
• Example added (in the documentation, to an existing notebook, or in a new notebook)
• Description in CHANGELOG.rst added (single line such as: (`#XX <https://github.com/iiasa/climate-assessment/pull/XX>`_) Added feature which does something)

[jkikstra]

Wrote an ad-hoc fix.

@phackstock if you need this soon, would you perhaps be able to take this a bit further (i.e. maybe a tests/docs, and check if my implementation makes sense)?

@znicholls FYI

[znicholls]

@znicholls FYI

Thanks, I put my comments in the issue to keep things in one place, implementation looks fine and I think adding tests/docs would be great (there's probably a much larger conversation to be had about whether we want to put more effort into cleaning things up in the repo in general, let me know if you have thoughts on that too).