km-kill-lags branched off frosting.

[kenmawer]

Closes issue #22.

[dajmcdon]

@kenmawer I'm not sure what the error is here, but this needs some debugging.

[kenmawer]

@dajmcdon These errors seem to be coming from tests of functions such as those relating to frosting, none of which I worked on. Here's an example from test-frosting:
Error in validate_is_function_set(mold): argument "mold" is missing, with no default

[dajmcdon]

Right, but those errors come from tests that use step_epi_lag(). So something that this branch does broke those tests. Two things I saw:

• Calling get_test_data() no longer works.
• In test-frosting.R on line 46, the predictions have the wrong date.

[dajmcdon]

Note that you can safely edit this branch locally and push to the same remote.

[dajmcdon]

If you merge frosting onto this branch, then the checks should pass.

[kenmawer]

As described earlier, splitting the print function means that the print won't work properly for lags.

[dajmcdon]

@kenmawer I fixed up the printing to work the way we discussed in my office. Why don't you pull locally, run the examples, and see what you think. Have a look at how I did it just to make sure that it makes sense. Once it does, you can approve this PR and merge merge this PR (I approve).