Melt per-vehicle-type availability from station_status into a long frame.
GBFS 2.2+/3.x stations may report vehicle_types_available (a list of
{vehicle_type_id, count}). This explodes it to one row per station × vehicle type
(STATION_VEHICLE_COUNTS_COLUMNS), so "where are the e-bikes?"
is a join to to_canonical_vehicle_types; the aggregate num_bikes_available
cannot answer it. Stations without the field contribute no rows.
See Also
to_canonical_station_status : The aggregate status parser.
to_canonical_vehicle_types : Resolve the vehicle types counted.
Source code in src/gbfs_toolkit/io/normalize.py
| def to_canonical_station_vehicle_counts(
raw: dict,
*,
system_id: str,
fetched_at: pd.Timestamp | None = None,
) -> pd.DataFrame:
"""Melt per-vehicle-type availability from ``station_status`` into a long frame.
GBFS 2.2+/3.x stations may report ``vehicle_types_available`` (a list of
``{vehicle_type_id, count}``). This explodes it to one row per station × vehicle type
(`STATION_VEHICLE_COUNTS_COLUMNS`), so "where are the e-bikes?"
is a join to [`to_canonical_vehicle_types`][gbfs_toolkit.to_canonical_vehicle_types]; the aggregate ``num_bikes_available``
cannot answer it. Stations without the field contribute no rows.
See Also
--------
[`to_canonical_station_status`][gbfs_toolkit.to_canonical_station_status] : The aggregate status parser.
[`to_canonical_vehicle_types`][gbfs_toolkit.to_canonical_vehicle_types] : Resolve the vehicle types counted.
"""
fetched_at = fetched_at if fetched_at is not None else _now_utc()
data = raw.get("data", raw)
stations = _records(data, "stations")
rows = [
{
"system_id": system_id,
"station_id": str(s.get("station_id")),
"vehicle_type_id": str(vt.get("vehicle_type_id")),
"num_vehicles_available": vt.get("count"),
"fetched_at": fetched_at,
}
for s in stations
for vt in (s.get("vehicle_types_available") or [])
]
df = pd.DataFrame(rows, columns=STATION_VEHICLE_COUNTS_COLUMNS)
df["num_vehicles_available"] = pd.to_numeric(
df["num_vehicles_available"], errors="coerce"
).astype("Int64")
df["vehicle_type_id"] = df["vehicle_type_id"].astype(
"string"
) # join key, see to_canonical_vehicles
df["fetched_at"] = pd.to_datetime(df["fetched_at"], utc=True)
return df
|