How long before the 6 May 2024 Osage County tornado was a tornado warning issued?

from datetime import UTC, datetime
from pathlib import Path

import matplotlib.dates as mdates
import matplotlib.pyplot as plt
import pandas as pd
from matplotlib.patches import Patch

import usdata
from usdata import cite_lockfile, pull, verify

# One figure style for every usdata notebook, so previews look alike.
plt.rcParams.update(
    {
        "figure.figsize": (8, 4.5),
        "figure.dpi": 120,
        "axes.spines.top": False,
        "axes.spines.right": False,
        "axes.grid": True,
        "grid.alpha": 0.3,
        "font.size": 10,
    }
)
manifest = Path("dataset.yaml")
print("Executed (UTC):", datetime.now(UTC).isoformat(timespec="seconds"))
print(f"usdata {usdata.__version__}; pandas {pd.__version__}")

print(manifest.read_text())

result = pull(manifest)
for item in result.fetched:
    print(item.asset.dataset_id)
    print(f"File: {item.path.name}")
    print(f"Bytes: {item.provenance.size}; cache hit: {item.from_cache}")
    print("Source:", item.provenance.source_url)
    print("Retrieved (UTC):", item.provenance.retrieved_at.isoformat(timespec="seconds"))
    print("Checksum:", item.provenance.checksum)
    print()

warnings = result.one("warnings").open_csv(parse_dates=["iso_issued", "iso_expired"])
reports = result.one("reports").open()
print(f"Storm Events rows for 2024: {len(reports):,}")
tornado = reports[reports["EVENT_ID"] == "1184254"].iloc[0]
start, end = tornado["BEGIN_UTC"], tornado["END_UTC"]
columns = ["EVENT_ID", "EVENT_TYPE", "TOR_F_SCALE", "CZ_NAME", "BEGIN_DATE_TIME"]
columns += ["CZ_TIMEZONE", "BEGIN_UTC", "END_UTC", "TOR_LENGTH", "TOR_OTHER_CZ_NAME"]
tornado[columns].to_frame("value")

print(f"{len(warnings)} events, from office(s): {sorted(warnings['wfo'].unique())}")
warnings.groupby(["ph_name", "sig_name"]).size().rename("events").to_frame()

tornado_warnings = warnings[
    (warnings["phenomena"] == "TO") & (warnings["significance"] == "W")
].sort_values("iso_issued")
minutes = pd.Timedelta(minutes=1)
table = pd.DataFrame(
    {
        "warning": tornado_warnings["eventid"],
        "issued (UTC)": tornado_warnings["iso_issued"].dt.strftime("%H:%M"),
        "expired (UTC)": tornado_warnings["iso_expired"].dt.strftime("%H:%M"),
        "issued, min before start": (start - tornado_warnings["iso_issued"]) // minutes,
        "expired, min after start": (tornado_warnings["iso_expired"] - start) // minutes,
        "in effect at start": (tornado_warnings["iso_issued"] <= start)
        & (tornado_warnings["iso_expired"] > start),
    }
).set_index("warning")
table

in_effect = tornado_warnings[
    (tornado_warnings["iso_issued"] <= start) & (tornado_warnings["iso_expired"] > start)
]
lead = start - in_effect["iso_issued"].min()

# Walk back from the start through overlapping tornado warnings to find how long
# some tornado warning naming the county had been continuously in effect.
covered_since = in_effect["iso_issued"].min()
for issued, expired in zip(
    tornado_warnings["iso_issued"][::-1], tornado_warnings["iso_expired"][::-1], strict=True
):
    if issued < covered_since <= expired:
        covered_since = issued
watch = warnings[(warnings["phenomena"] == "TO") & (warnings["significance"] == "A")].iloc[0]

print(f"The EF4 began at {start:%H:%M} UTC on {start:%d %B %Y}.")
print(
    f"Tornado Warning {in_effect['eventid'].iloc[0]} was in effect, "
    f"issued {lead // minutes} minutes earlier."
)
print(
    f"A tornado warning naming the county had been in effect continuously since "
    f"{covered_since:%H:%M} UTC, {(start - covered_since) // minutes} minutes before."
)
print(
    f"The Tornado Watch was issued at {watch['iso_issued']:%H:%M} UTC, "
    f"{(start - watch['iso_issued']) / pd.Timedelta(hours=1):.1f} hours before."
)

window_start = pd.Timestamp("2024-05-07T01:00Z")
window_end = pd.Timestamp("2024-05-07T04:00Z")
shown = warnings[
    (warnings["iso_expired"] > window_start) & (warnings["iso_issued"] < window_end)
].sort_values("iso_issued", ascending=False)
colors = {("TO", "W"): "#b42318", ("SV", "W"): "#d99a1e", ("TO", "A"): "#e8b4ae"}

fig, ax = plt.subplots(layout="constrained")
ax.axvspan(mdates.date2num(start), mdates.date2num(end), color="#444444", alpha=0.12)
ax.axvline(mdates.date2num(start), color="#444444", linewidth=0.9)
for row, event in enumerate(shown.itertuples()):
    left = mdates.date2num(max(event.iso_issued, window_start))
    right = mdates.date2num(min(event.iso_expired, window_end))
    color = colors.get((event.phenomena, event.significance), "#9aa0a6")
    ax.barh(row, right - left, left=left, height=0.6, color=color)
ax.set_yticks(range(len(shown)))
ax.set_yticklabels(f"{e.ph_name} {e.sig_name} {e.eventid}" for e in shown.itertuples())
ax.set_xlim(mdates.date2num(window_start), mdates.date2num(window_end))
ax.xaxis.set_major_locator(mdates.MinuteLocator(byminute=[0, 30], tz=UTC))
ax.xaxis.set_major_formatter(mdates.DateFormatter("%H:%M", tz=UTC))
ax.set_xlabel("7 May 2024 (UTC)")
ax.grid(axis="y", visible=False)
ax.set_title("Osage County, OK: NWS alerts around the 6 May 2024 EF4")
fig.legend(
    handles=[
        Patch(color=colors[("TO", "W")], label="Tornado Warning"),
        Patch(color=colors[("SV", "W")], label="Severe Thunderstorm Warning"),
        Patch(color=colors[("TO", "A")], label="Tornado Watch"),
        Patch(color="#444444", alpha=0.12, label="EF4 on the ground (Storm Events)"),
    ],
    loc="outside lower center",
    ncols=4,
    frameon=False,
    fontsize=8,
)
plt.show()

assert verify(manifest) == []
for citation in cite_lockfile(manifest):
    print(citation.as_text())