v1.103.0

Added

• Optional default= argument on ResultFloat and ResultVec (defaults to 0, so no behaviour change). The hardcoded 0 default meant an unrecorded sample — a run that aborts before measuring, or a result var the worker never sets — was indistinguishable from a real 0 measurement, dragging nan-aware regression/aggregation means toward zero. Callers can now opt in with default=float("nan") so unrecorded samples are treated as missing and dropped by the existing np.nanmean/np.nansum reductions. default is not a hashed slot, so opting in does not invalidate over_time history for an otherwise-identical result var.
• test/test_result_nan_default.py: backward-compat (default still 0), NaN/explicit-default opt-in, hash stability, end-to-end unrecorded-sample handling, plus serialization coverage — a pickle save_result/load_result round-trip and a HoloViews→bokeh render_report HTML render both preserve/handle the NaN default.