Files
aimsight-labeling/auto/analyses/long-labels/FLORIDA/analysis.md
T
maximeandClaude Opus 5.5 ddb17c2189 auto: labels requalified under the 5-noise rules (only what stands out from the noise): 32 analyses refined, the 6 paused periods verified, the long labels requalified with the fixes the period analysts found; 1153 -> 961 labels
- 32 analyses: each label kept, rewritten or removed by its analyst (section Requalification of the report; mm magnitudes where the targets have them)
- SNCF01 x3, WILDSTRUBEL 2026-09-01 -> 2026-10-06 x3: verified from their existing labels under the new rules
- long labels (no analysis): auto/analyses/long-labels/<STATION>/; NORDEND-169 and -195 run to 07/11, so NORDEND-283 and -284 (their continuation) are removed
- reading-rules.md, stations.md: the citations of removed or rewritten labels point to their successor, or to [expert]

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-17 23:58:00 +01:00

1.4 KiB
Raw Blame History

Requalification of the long labels: FLORIDA, 2026-10-01 → 2026-11-01 (UTC)

Evidence: work/inventory.txt (rejections over the window), work/ranges_k5.txt (ranges.py --k 5, corr_x, corr_y and their mm, accepted measurements only), work/ranges_k5_roll1d.txt (same with --rolling 1d).

  • FLORIDA-1 (Sellette 6, Target issue, 10-01 → 11-01): kept as is. 1651 of about 2976 measurements rejected over the month (about 55 %), own runs of rejections every day, day and night. Its accepted measurements leave the noise: corr_x up to 30.8 noises (+0.94 px, +14.9 mm, 10-23 14:30), corr_y up to −29 noises (−0.69 px, −10.9 mm, 10-01 21:15), 11 ranges of ≥ 6 measurements beyond 5 noises on corr_x and 20 on corr_y, spread over the month. Constant problem target whose kept points jump beyond the noise: the label holds.
  • FLORIDA-14 (Massif 2, Target issue, 10-06 10:15 → 11-01): kept as is. About 1430 rejections after 10-06 10:15 out of about 2490 measurements (about 57 %; majority rejected day and night from the 8th to the 29th, runs of 55 to 267 consecutive rejections). Its accepted measurements leave the noise: corr_y down to −0.95 px (−18.7 mm, −29.7 noises), level shift of corr_y over 10-30 → 11-01 (118 measurements in a row beyond −5 noises, down to −21 noises, −13.4 mm), corr_x jumps up to 36.6 noises (−1.04 px, −20.4 mm). Degrading target whose kept points change level beyond the noise: the label holds.