- 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>
78 lines
3.5 KiB
Python
78 lines
3.5 KiB
Python
"""Requalification of the labels of the analysis: for each label, runs tools/scripts/ranges.py at --k 5 on corr_x, corr_y
|
|
and keeps the ranges of the targets (global references CTR02, CTR03, CTR04 excluded) that fall inside the label.
|
|
|
|
- movement labels (Visibility, Displacement, Obstruction): window = label +- 3 h, --outside label (level and noise
|
|
of the hours around); labels longer than 4 h: window = the label, --rolling 8 (local level under the thermal cycle).
|
|
- Thermal displacement: window = 5 h before the start (end of night) + label, --outside label: level and noise of the
|
|
night that precedes; instants covered by a non-thermal label are not counted; ranges of >= 3 measurements reported.
|
|
- No data, and the labels of missing points: their ranges are listed too (what the other curves do meanwhile).
|
|
Usage: python3 <this> <labels.jsonl> <out.txt>
|
|
"""
|
|
import json, re, subprocess, sys
|
|
from datetime import datetime, timedelta, timezone
|
|
|
|
REFS = {"CTR02", "CTR03", "CTR04"}
|
|
F = "%Y-%m-%dT%H:%MZ"
|
|
|
|
|
|
def p(s):
|
|
return datetime.strptime(s, F).replace(tzinfo=timezone.utc)
|
|
|
|
|
|
def f(t):
|
|
return t.strftime(F)
|
|
|
|
|
|
def ranges(start, end, extra):
|
|
cmd = ["python3", "tools/scripts/ranges.py", "data/NORDEND", "--columns", "corr_x,corr_y", "--k", "5",
|
|
"--start", f(start), "--end", f(end)] + extra
|
|
out = subprocess.run(cmd, capture_output=True, text=True).stdout
|
|
res, col = [], None
|
|
for line in out.splitlines():
|
|
m = re.match(r"(corr_[xy]): ranges more than", line)
|
|
if m:
|
|
col = m.group(1)
|
|
continue
|
|
if line.startswith("=="):
|
|
col = None
|
|
if col:
|
|
tok = line.split()
|
|
if len(tok) >= 7 and re.match(r"\d{4}-", tok[1]):
|
|
res.append(dict(col=col, s=tok[0], a=p(tok[1]), b=p(tok[2]), n=int(tok[3]), v=float(tok[4]),
|
|
k=float(tok[5]), at=tok[6]))
|
|
return cmd, res
|
|
|
|
|
|
labels = [json.loads(l) for l in open(sys.argv[1])]
|
|
other = [(p(l["start"]), p(l["end"])) for l in labels if l["family"] not in ("Thermal displacement", "No data")]
|
|
out = open(sys.argv[2], "w")
|
|
for l in labels:
|
|
s, e = p(l["start"]), p(l["end"])
|
|
fam = l["family"]
|
|
if fam == "Thermal displacement":
|
|
cmd, rs = ranges(s - timedelta(hours=5), e, ["--outside", f"{f(s)}/{f(e)}"])
|
|
keep = []
|
|
for r in rs:
|
|
if r["s"] in REFS or r["b"] <= s or r["a"] >= e:
|
|
continue
|
|
# measurements of the range outside the non-thermal labels: approximated by the range span
|
|
inside_other = any(a < r["a"] and r["b"] < b for a, b in other)
|
|
if not inside_other and r["n"] >= 3:
|
|
keep.append(r)
|
|
else:
|
|
long = e - s > timedelta(hours=4)
|
|
if long:
|
|
cmd, rs = ranges(s, e, ["--rolling", "8"])
|
|
else:
|
|
cmd, rs = ranges(s - timedelta(hours=3), e + timedelta(hours=3), ["--outside", f"{f(s)}/{f(e)}"])
|
|
keep = [r for r in rs if r["s"] not in REFS and r["a"] < e and r["b"] > s]
|
|
keep.sort(key=lambda r: -abs(r["k"]))
|
|
best = keep[0] if keep else None
|
|
out.write(f"{l.get('id')} {l['start']} {l['end']} {fam}\n cmd: {' '.join(cmd[1:])}\n")
|
|
out.write(f" max: {best['s']} {best['col']} {best['k']:+.1f} noises ({best['v']:+.4f}) at {best['at']} n={best['n']}\n"
|
|
if best else " max: none beyond 5 noises\n")
|
|
for r in keep[:8]:
|
|
out.write(f" {r['s']:6} {r['col']} {f(r['a'])}..{f(r['b'])} n={r['n']} {r['k']:+.1f} at {r['at']}\n")
|
|
out.write("\n")
|
|
out.flush()
|