[us-commits] [ehb54/ultrascan3] c9ac11: somo/doc: document the GRPY Numerical Precision co...

[emre brookes]

  Branch: refs/heads/ehb54-issue-984
  Home:   https://github.com/ehb54/ultrascan3
  Commit: c9ac1177ffb801eaf2ec54bf734b72f5329a554c
      https://github.com/ehb54/ultrascan3/commit/c9ac1177ffb801eaf2ec54bf734b72f5329a554c
  Author: ehb54 <brookes at uthscsa.edu>
  Date:   2026-08-06 (Thu, 06 Aug 2026)

  Changed paths:
    M us_somo/somo/doc/manual/somo/somo_hydro.html

  Log Message:
  -----------
  somo/doc: document the GRPY Numerical Precision control

The Double/Float control shipped with the in-process GRPY work but was never
documented -- a grep across the whole manual returned nothing for it, despite
the box sitting directly above the shell-reduction one just added. Same file,
same window, so it is described here in window order: buried-bead exclusion,
precision, shell reduction.

Explains what the setting actually governs (the precision in which the large
internal matrix is stored and factored), why it matters (that matrix holds
eleven quantities per bead and grows as the square of the bead count, so it
is what limits the treatable model size), and that the tensor arithmetic and
all subsequent processing stay in double precision regardless.

States the accuracy cost from validation rather than in the abstract: every
labelled observable is identical to the four significant figures displayed,
with only the near-zero reoriented coupling terms differing. Recommends
Double as the default nonetheless, since the benefit appears only when memory
is the binding constraint.

Fixes ehb54/ultrascan-tickets#984

Co-Authored-By: Claude Opus 5 <noreply at anthropic.com>



To unsubscribe from these emails, change your notification settings at https://github.com/ehb54/ultrascan3/settings/notifications