Branch: refs/heads/ehb54-issue-920 Home: https://github.com/ehb54/ultrascan3 Commit: e2d0c3b012a50fefb15cc2759bd2284830f4e185 https://github.com/ehb54/ultrascan3/commit/e2d0c3b012a50fefb15cc2759bd228483... Author: ehb54 <brookes@uthscsa.edu> Date: 2026-07-13 (Mon, 13 Jul 2026) Changed paths: M us_somo/develop/src/us_hydrodyn_saxs_iqq_extrap_c0.cpp Log Message: ----------- extrap_c0: scale the spliced reference for a continuous merge cutover The high-q reference splice copied the reference verbatim, but the dilute extrapolation and the reference generally differ by a few percent at the merge point (the extrapolation carries the true dilute-limit scale), leaving a visibly bad step at the cutover (~5% on the alpha-syn test set, obvious near q=0.09). Scale the spliced reference by the median ratio of the normalized extrapolation to the normalized reference over a 25-point window just below the merge, so the reference supplies clean high-q shape stitched smoothly onto the extrapolation (the way almerge's scaled merge stays continuous). Removes the systematic step, leaving only noise-level mismatch; the factor is reported in the editor. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> To unsubscribe from these emails, change your notification settings at https://github.com/ehb54/ultrascan3/settings/notifications