app.api.dashboard

The dashboard’s REST API — the calls frontend/js/api.js makes, in the shapes mock.js defines.

Functions

vehicles_now(state)

Classes

WhatifRequest(*args, **kwargs)

class app.api.dashboard.WhatifRequest(*args: Any, **kwargs: Any)

Базовые классы: BaseModel

at_stop_id: str | None = None

route_id: str

scenario: Literal['add_reserve', 'adjust_interval', 'detour', 'signal_priority', 'hold_at_stop']

app.api.dashboard.metrics_bunching() → list[dict]

Always empty: bunching needs vehicles sharing a route and direction, and the dataset has no route relations between vehicles (each vehicle is its own «route»).

async app.api.dashboard.metrics_model(request: fastapi.Request) → dict

The ML service’s model metrics (cached; last known copy while ML is down) plus live backend metrics.

app.api.dashboard.metrics_worst_stops(request: fastapi.Request, limit: int = fastapi.Query) → list[dict]

Stops with the largest forecast delays over the last 30 minutes (average ≥ 30 s).

app.api.dashboard.route_signals(route_id: str) → list[dict]

Always empty: the dataset has no traffic-light data (the dashboard’s signal layer is a mock-mode feature).

app.api.dashboard.routes(request: fastapi.Request) → list[dict]

One «route» per scheduled vehicle (the dataset has no route ids); each distinct trip shape is a direction.

app.api.dashboard.schedule(request: fastapi.Request, vehicle_id: int) → dict

The vehicle’s current trip, stop by stop: actual arrivals behind it, forecasts ahead.

app.api.dashboard.vehicles(request: fastapi.Request) → list[dict]

Vehicles in service: position, current deviation, the forecast for their target stop and its risk.

app.api.dashboard.vehicles_now(state) → list[dict]

async app.api.dashboard.whatif(request: fastapi.Request, body: WhatifRequest) → dict

Scenario effect on the vehicle’s latest forecast, from ML /whatif/predict; also pushed as whatif.result.