There may be edge cases or other variations on this we want to add
later, but this makes sure that we at least have some coverage of the
method.
Closes #110
Note to Reviewer
I was initially a little surprised that the CSV generated in the second case had a name header, as the test data only uses id. But this is because the subjectdefined at line 4 explicitly sets the columns for both groups of tests. I suspect that each group should really have its own subject defined, but changing that seems like something that should really happen in a different PR.
There may be edge cases or other variations on this we want to add later, but this makes sure that we at least have some coverage of the method.
Closes #110
Note to Reviewer
I was initially a little surprised that the CSV generated in the second case had a
nameheader, as the test data only usesid. But this is because thesubjectdefined at line 4 explicitly sets the columns for both groups of tests. I suspect that each group should really have its ownsubjectdefined, but changing that seems like something that should really happen in a different PR.