app.api.views

Dashboard views: routes, vehicles and schedules in the shapes frontend/js/mock.js defines.

The dataset has no routes — only each vehicle’s planned stop visits for the day. So route_id is the vehicle id, and every distinct trip shape (grouped by its first and last stop) is one «direction» with its own line: the map projects the vehicle and its schedule onto the line of the vehicle’s current trip, which a back-and-forth whole-day polyline would make ambiguous. Lines are stop-to-stop straight segments (there is no road geometry in the dataset).

Functions

current_prediction(p, T)	The latest forecast, while its target stop is still ahead.
next_pos(v, T)	Next planned visit: after the last detected arrival if that's recent, otherwise by plan time and current deviation.
schedule_payload(state, v, T)	The vehicle's current trip: passed stops with the detected arrival, the target stop with the model's forecast, other upcoming stops with the current deviation carried forward.
vehicle_payload(state, v, T, max_age_s)
worst_stops(state, T, limit[, window_s])	Stops with the largest forecast delays over the last window_s of dataset time.

Classes

RouteCatalog(schedules)

Built once from the schedules: route payloads and the direction of every trip.

class app.api.views.RouteCatalog(schedules: dict[int, VehicleSchedule])

Базовые классы: object

Built once from the schedules: route payloads and the direction of every trip.

__init__(schedules: dict[int, VehicleSchedule]) → None

direction(tr_id: int, trip: int) → int

app.api.views.current_prediction(p: Prediction | None, T: datetime) → Prediction | None

The latest forecast, while its target stop is still ahead.

app.api.views.next_pos(v: VehicleState, T: datetime) → int

Next planned visit: after the last detected arrival if that’s recent, otherwise by plan time and current deviation.

app.api.views.schedule_payload(state, v: VehicleState, T: datetime) → dict

The vehicle’s current trip: passed stops with the detected arrival, the target stop with the model’s forecast, other upcoming stops with the current deviation carried forward.

app.api.views.vehicle_payload(state, v: VehicleState, T: datetime, max_age_s: float) → dict | None

app.api.views.worst_stops(state, T: datetime, limit: int, window_s: float = 1800) → list[dict]

Stops with the largest forecast delays over the last window_s of dataset time.