Add multi-index support to d1s.apply_time_correction (sequence in, list out)

[shimwell]

Summary

apply_time_correction now takes a sequence of time indices and returns a list of derived tallies, one per index. The tally arrays are read and reshaped once and shared across all indices, so post-processing a D1S mesh tally at many shutdown times no longer re-reads and re-reshapes the data on every call.

The per-index math is unchanged from develop, so each derived tally is bit-for-bit identical to the previous single-index result.

API change (small, intentional break)

This favours a clean, type-stable signature over a polymorphic int | Sequence[int] → Tally | list[Tally] API:

• index is now an iterable of ints; the default is (-1,) (was -1).
• The function always returns list[openmc.Tally], one per index, in order.

Callers that want a single time unpack the one-element list:

tally, = d1s.apply_time_correction(dose_tally, factors, [-1])

The existing test, the user-guide example, and a new multi-index test are updated accordingly. A .. versionchanged:: 0.16.0 note documents the break in the docstring.

Testing

Adds test_apply_time_correction_multi_index, which checks that each element of a multi-index call matches the corresponding single-index call (mean, std_dev, filters, and sum/sum_sq), that unordered/partial index sequences are honored in order, and that the original tally is left unmodified. The existing test_apply_time_correction is updated to unpack the list.

Note: the D1S unit tests call model.run() and require a compiled OpenMC plus nuclear data, so they were not executed locally — CI will cover them.

Checklist

• I have performed a self-review of my own code
• I have run clang-format (version 18) on any C++ source files (if applicable)
• I have followed the style guidelines for Python source files (if applicable)
• I have made corresponding changes to the documentation (if applicable)
• I have added tests that prove my fix is effective or that my feature works (if applicable)