Clarify elem comment in jump_error_estimator

[lindsayad]

In jump_error_estimator.C

[roystgnr]

If we have 2 elems with ids 10 and 11, we want a vector that will take indices 10 and 11. A vector of length 2 won't cut it.

What prompted this PR?

[moosebuild]

Job Coverage, step Generate coverage on 15ab5f1 wanted to post the following:

Coverage

	61acf4	#4356 15ab5f
	Total	Total	+/-	New
Rate	65.27%	65.27%	-0.00%	-
Hits	77399	77398	-1	0
Misses	41185	41186	+1	0

Diff coverage report

Full coverage report

This comment will be updated on new commits.

[lindsayad]

If we have 2 elems with ids 10 and 11, we want a vector that will take indices 10 and 11. A vector of length 2 won't cut it.

What prompted this PR?

I read the comment and immediately surrounding code and figured that "resizing error_per_cell vector to be the number of elements" would be a nice way to sync the comment and the code. I can see later on in the code that we index with ids so indeed this isn't a good change. But the fact that there are no issues in CIVET suggest there is no coverage of this code?

  // Resize the error_per_cell vector to be
  // the number of elements, initialize it to 0.
  error_per_cell.resize (mesh.n_elem());
  std::fill (error_per_cell.begin(), error_per_cell.end(), 0.);

[roystgnr]

We've probably got line coverage but not feature coverage - we'd need to do adaptive coarsening while disabling renumbering to trigger a bug here, I think.

[lindsayad]

K I just changed the comment instead