File size: 9,275 Bytes
b4a9114 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 | #!/usr/bin/env python3
"""Calculate precision proxy 1 and its reference-window sensitivity.
The script extracts one declared output from the compact multi-output catalogue,
intersects merged manual-reference windows with that output's eligibility
domain, and performs deterministic one-to-one station-phase-time matching.
"""
from __future__ import annotations
import argparse
import csv
import json
import sqlite3
import sys
from collections import defaultdict
from pathlib import Path
from typing import Iterable
import numpy as np
ROOT = Path(__file__).resolve().parents[1]
if str(ROOT) not in sys.path:
sys.path.insert(0, str(ROOT))
from scripts.validate_leave_one_output_precision_proxy import (
PERIODS,
assign_sorted_one_to_one,
iter_station_phase_groups,
load_eligibility,
)
def load_manual_references(
path: Path,
) -> dict[tuple[str, str, str], np.ndarray]:
con = sqlite3.connect(path)
grouped: dict[tuple[str, str, str], list[int]] = defaultdict(list)
for period, station, phase, epoch in con.execute(
"""
SELECT period, station_id, phase, pick_time_epoch
FROM evaluation_references
WHERE has_manual_reference=1 AND c0_point_covered=1
ORDER BY period, station_id, phase, pick_time_epoch
"""
):
grouped[(str(period), str(station), str(phase))].append(
int(round(float(epoch) * 1_000_000))
)
con.close()
return {key: np.asarray(values, dtype=np.int64) for key, values in grouped.items()}
def merge_windows(times: np.ndarray, half_width_us: int) -> list[tuple[int, int]]:
merged: list[tuple[int, int]] = []
for value in times:
left = int(value) - half_width_us
right = int(value) + half_width_us
if merged and left <= merged[-1][1]:
merged[-1] = (merged[-1][0], max(merged[-1][1], right))
else:
merged.append((left, right))
return merged
def intersect_windows(
left: list[tuple[int, int]],
right_starts: np.ndarray,
right_ends: np.ndarray,
) -> list[tuple[int, int]]:
result: list[tuple[int, int]] = []
i = 0
j = 0
while i < len(left) and j < len(right_starts):
start = max(left[i][0], int(right_starts[j]))
end = min(left[i][1], int(right_ends[j]))
if start <= end:
if result and start <= result[-1][1]:
result[-1] = (result[-1][0], max(result[-1][1], end))
else:
result.append((start, end))
if left[i][1] <= right_ends[j]:
i += 1
else:
j += 1
return result
def select_in_windows(times: np.ndarray, windows: list[tuple[int, int]]) -> np.ndarray:
selected: list[np.ndarray] = []
for left, right in windows:
begin = int(np.searchsorted(times, left, side="left"))
end = int(np.searchsorted(times, right, side="right"))
if end > begin:
selected.append(times[begin:end])
return np.concatenate(selected) if selected else np.empty(0, dtype=np.int64)
def main() -> None:
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument(
"--candidates", type=Path,
default=ROOT / "data/label/all_model_phase_picks.jsonl.zst",
)
parser.add_argument(
"--references", type=Path,
default=ROOT / "data/label/reference_arrivals.sqlite",
)
parser.add_argument(
"--eligibility", type=Path,
default=ROOT / "data/label/output_eligibility.sqlite",
)
parser.add_argument("--output-name", default="seismicxm.01")
parser.add_argument("--epsilon-s", type=float, default=0.5)
parser.add_argument("--window-s", type=float, action="append")
parser.add_argument(
"--output-tsv", type=Path,
default=ROOT / "data/validation/reference_window_precision_proxy_sensitivity.tsv",
)
parser.add_argument(
"--output-json", type=Path,
default=ROOT / "data/validation/reference_window_precision_proxy_sensitivity.json",
)
args = parser.parse_args()
windows_s = args.window_s or [1.0, 5.0, 10.0, 30.0]
if any(value < args.epsilon_s for value in windows_s):
raise ValueError("every T must be at least epsilon")
references = load_manual_references(args.references)
reference_totals: dict[tuple[str, str], int] = defaultdict(int)
for (period, _, phase), values in references.items():
reference_totals[(period, phase)] += len(values)
eligibility, output_names = load_eligibility(args.eligibility)
name_to_id = {name: output_id for output_id, name in output_names.items()}
if args.output_name not in name_to_id:
raise KeyError(f"output is absent from registry: {args.output_name}")
output_id = name_to_id[args.output_name]
epsilon_us = int(round(args.epsilon_s * 1_000_000))
counters: dict[tuple[str, str, float], list[int]] = defaultdict(
lambda: [0, 0, 0, 0, 0]
)
def process_group(station: str, phase: str, assessed: np.ndarray) -> None:
domain = eligibility.get((station, phase, output_id))
if domain is None:
return
domain_starts, domain_ends = domain
for period in PERIODS:
refs = references.get((period, station, phase))
if refs is None or not len(refs):
continue
ref_indices = np.searchsorted(domain_starts, refs, side="right") - 1
valid = ref_indices >= 0
valid &= domain_ends[np.maximum(ref_indices, 0)] >= refs
eligible_refs = refs[valid]
if not len(eligible_refs):
continue
for window_s in windows_s:
raw_windows = merge_windows(
eligible_refs, int(round(window_s * 1_000_000))
)
effective = intersect_windows(raw_windows, domain_starts, domain_ends)
selected_picks = select_in_windows(assessed, effective)
assignments = assign_sorted_one_to_one(
eligible_refs, selected_picks, epsilon_us
)
matched = int(np.sum(assignments >= 0))
values = counters[(period, phase, float(window_s))]
values[0] += len(refs)
values[1] += len(eligible_refs)
values[2] += len(effective)
values[3] += int(sum(right - left for left, right in effective))
values[4] += len(selected_picks)
if len(values) == 5:
values.append(0)
values[5] += matched
seen_groups: set[tuple[str, str]] = set()
for group_index, (station, phase, times, models) in enumerate(
iter_station_phase_groups(args.candidates), 1
):
seen_groups.add((station, phase))
assessed = times[models == output_id]
process_group(station, phase, assessed)
if group_index % 100 == 0:
print(f"[PROXY1] groups={group_index:,}", flush=True)
for _, station, phase in references:
if (station, phase) not in seen_groups:
process_group(station, phase, np.empty(0, dtype=np.int64))
seen_groups.add((station, phase))
rows = []
for (period, phase, window_s), values in sorted(counters.items()):
_, eligible_refs, n_windows, duration_us, picks, matched = values
rows.append(
{
"output_id": output_id,
"output_name": args.output_name,
"period": period,
"phase": phase,
"window_half_width_T_s": window_s,
"matching_tolerance_epsilon_s": args.epsilon_s,
"c0_manual_reference_count": reference_totals[(period, phase)],
"workflow_eligible_reference_count": eligible_refs,
"effective_window_count": n_windows,
"effective_window_duration_s": duration_us / 1_000_000,
"reported_pick_count_in_effective_windows": picks,
"matched_pick_count": matched,
"catalogue_conditioned_precision_proxy": matched / picks if picks else None,
}
)
payload = {
"schema": "seismicx-cont-reference-window-precision-proxy-v1",
"candidate_file": args.candidates.relative_to(ROOT).as_posix(),
"reference_file": args.references.relative_to(ROOT).as_posix(),
"eligibility_file": args.eligibility.relative_to(ROOT).as_posix(),
"reference_definition": (
"canonical C0-covered station-location-phase-time references with "
"at least one manual source reading"
),
"assignment": "deterministic earliest-feasible one-to-one station-phase-time matching",
"rows": rows,
}
args.output_json.parent.mkdir(parents=True, exist_ok=True)
args.output_json.write_text(json.dumps(payload, indent=2, sort_keys=True) + "\n")
with args.output_tsv.open("w", newline="") as stream:
writer = csv.DictWriter(stream, fieldnames=list(rows[0]), delimiter="\t")
writer.writeheader()
writer.writerows(rows)
print(json.dumps({"rows": len(rows), "output": args.output_name}, indent=2))
if __name__ == "__main__":
main()
|