66 lines
3.4 KiB
Python
66 lines
3.4 KiB
Python
"""Every target measurement more than K own noises from its local level, grouped by instant, against the labels.
|
|||
|
|
|
||
|
|
Noise: median |diff| of consecutive accepted measurements over the range (as ranges.py).
|
||
|
|
Local level: median of the N accepted neighbours on each side, itself excluded (as ranges.py --rolling N).
|
||
|
|
Output 1: per instant with at least one target > K: the targets and their signed deviation (noises), the
|
||
|
|
references out (rejected r / inlier 0), and the labels covering it (start < t < end).
|
||
|
|
Output 2: per label: the max deviation strictly inside, with its target and instant; '-' if none > K.
|
||
|
|
Usage: python3 exceed.py LABELS.jsonl [--k 5] [--n 8] [--uncovered]
|
||
|
|
"""
|
||
|
|
import json, sys, statistics
|
||
|
|
sys.path.insert(0, "tools/scripts")
|
||
|
|
import data
|
||
|
|
|
||
|
|
K = float(sys.argv[sys.argv.index("--k") + 1]) if "--k" in sys.argv else 5
|
||
|
|
N = int(sys.argv[sys.argv.index("--n") + 1]) if "--n" in sys.argv else 8
|
||
|
|
S, E = data.parse_time("2026-09-24T02:30Z"), data.parse_time("2026-10-06T19:45Z")
|
||
|
|
by = data.load("data/WILDSTRUBEL")
|
||
|
|
for _ms in by.values(): # measurement instants to the minute, as the label bounds
|
||
|
|
for _m in _ms:
|
||
|
|
_m["t"] = _m["t"].replace(second=0, microsecond=0)
|
||
|
|
refs = sorted(s for s, ms in by.items() if data.is_reference(ms))
|
||
|
|
tg = sorted(s for s, ms in by.items() if not data.is_reference(ms))
|
||
|
|
labels = [json.loads(l) for l in open(sys.argv[1])]
|
||
|
|
for L in labels:
|
||
|
|
L["a"], L["b"] = data.parse_time(L["start"]), data.parse_time(L["end"])
|
||
|
|
|
||
|
|
dev = {} # t -> list of (target, col, signed noises)
|
||
|
|
for s in tg:
|
||
|
|
for c in ("corr_x", "corr_y"):
|
||
|
|
acc = [m for m in by[s] if m[c] is not None and S <= m["t"] <= E]
|
||
|
|
v = [m[c] for m in acc]
|
||
|
|
nz = statistics.median(abs(b - a) for a, b in zip(v, v[1:]))
|
||
|
|
for i, m in enumerate(acc):
|
||
|
|
nb = v[max(0, i - N):i] + v[i + 1:i + 1 + N]
|
||
|
|
d = (v[i] - statistics.median(nb)) / nz
|
||
|
|
dev.setdefault(m["t"], []).append((s, c[-1], d))
|
||
|
|
|
||
|
|
refout = {}
|
||
|
|
for r in refs:
|
||
|
|
for m in by[r]:
|
||
|
|
if S <= m["t"] <= E and (data.rejected(m) or m.get("inlier") == 0):
|
||
|
|
refout.setdefault(m["t"], []).append(r + ("r" if data.rejected(m) else "0"))
|
||
|
|
|
||
|
|
def cover(t):
|
||
|
|
return [L.get("id", "new").replace("WILDSTRUBEL-", "") for L in labels if L["a"] < t < L["b"]]
|
||
|
|
|
||
|
|
if "--per-label" not in sys.argv:
|
||
|
|
for t in sorted(dev):
|
||
|
|
over = [(s, c, d) for s, c, d in dev[t] if abs(d) > K]
|
||
|
|
if not over:
|
||
|
|
continue
|
||
|
|
cv = cover(t)
|
||
|
|
if "--uncovered" in sys.argv and [x for x in cv if x not in sys.argv[sys.argv.index("--uncovered") + 1].split(",")]:
|
||
|
|
continue
|
||
|
|
over.sort(key=lambda x: -abs(x[2]))
|
||
|
|
print(f"{t:%m-%d %H:%M} " + " ".join(f"{s}.{c}{d:+.1f}" for s, c, d in over)
|
||
|
|
+ f" | refs out: {','.join(x for x in refout.get(t, []) if x != 'R3r') or '-'} | labels: {','.join(cv) or 'NONE'}")
|
||
|
|
print("\n== per label: max |dev| strictly inside (rolling), n instants > K ==")
|
||
|
|
for L in labels:
|
||
|
|
ins = [(abs(d), s, c, d, t) for t in dev if L["a"] < t < L["b"] for s, c, d in dev[t]]
|
||
|
|
if not ins:
|
||
|
|
print(f"{L.get('id','new'):16s} {L['start'][5:16]} {L['end'][5:16]} {L['family'][:10]:10s} no measurement inside")
|
||
|
|
continue
|
||
|
|
mx = max(ins)
|
||
|
|
n = len({t for a, s, c, d, t in ins if a > K})
|
||
|
|
print(f"{L.get('id','new'):16s} {L['start'][5:16]} {L['end'][5:16]} {L['family'][:10]:10s} max {mx[1]}.{mx[2]} {mx[3]:+.1f} @{mx[4]:%m-%d %H:%M} instants>K {n}")
|