table: fix dt-mean

[oliver-sanders]

Applies the fix from #1112 onto master and adds tests.

Requirements check-list

• I have read CONTRIBUTING.md and added my name as a Code Contributor.
• Contains logically grouped changes (else tidy your branch by rebase).
• Does not contain off-topic changes (use other PRs for other changes).
• Appropriate tests are included (unit and/or functional).
• Appropriate change log entry included.
• No documentation update required.

[MetRonnie]

What about the delta symbol instead of "d"?

[oliver-sanders]

What about the delta symbol instead of "d"?

Could do but I'm not sure that was a positive change. Should probably rethink the timing headers some time.

[MetRonnie]

Oh, test failing after my suggestion was committed because it was expecting a ΔT of 0 to be converted to ''. I have had a look at a real workflow and it seems that runahead tasks that have not run in a previous cycle get a mean elapsed time of 0 when it should probably be nullish (running tasks that have not run in a previous cycle get a nullish value). In the mock data there are several tasks that have a mean elapsed time of 0 which should probably be >0 if the task has a T-finish, else nullish.

Or, you could just undo my suggestion

[oliver-sanders]

Going to revert the suggestion in order to keep this as robust as possible.

[MetRonnie]

1 suggestion for clarifying comment

[wxtim]

@oliver-sanders - this now conflicts with master after merger of #1123

[wxtim]

I've just tried this out and it doesn't seem to be working at all for me.

[oliver-sanders]

Rebased

[MetRonnie]

I have tested it out and it's working for me. Just the changelog that is conflicting

[wxtim]

• Checks that it works in practice.

[wxtim]

• Read code.