@spec api_web_prediction_controller_index(
  Tesla.Env.client(),
  keyword()
) ::
  {:ok, MBTA.Model.Forbidden.t()}
  | {:ok, MBTA.Model.TooManyRequests.t()}
  | {:ok, MBTA.Model.Predictions.t()}
  | {:ok, MBTA.Model.BadRequest.t()}
  | {:error, Tesla.Env.t()}

NOTE: A filter MUST be present for any predictions to be returned. List of predictions for trips. To get the scheduled times instead of the predictions, use /schedules. The predicted arrival time (//data/{index}/attributes/arrival_time) and departure time (/data/{index}/attributes/departure_time) to/from a stop (/data/{index}/relationships/stop/data/id) at a given sequence (/data/{index}/attriutes/stop_sequence) along a trip (/data/{index}/relationships/trip/data/id) going a direction (/data/{index}/attributes/direction_id) along a route (/data/{index}/relationships/route/data/id). See GTFS Realtime FeedMesage FeedEntity TripUpdate TripDescriptor See GTFS Realtime FeedMesage FeedEntity TripUpdate StopTimeUpdate ## When a vehicle is predicted to be at a stop /predictions?filter[stop]=STOP_ID ## The predicted schedule for one route /predictions?filter[route]=ROUTE_ID ## The predicted schedule for a whole trip /predictions?filter[trip]=TRIP_ID

Parameters

• connection (MBTA.Connection): Connection to server
• opts (keyword): Optional parameters
  • :"page[offset]" (integer()): Offset (0-based) of first element in the page
  • :"page[limit]" (integer()): Max number of elements to return

  • :sort (String.t): Results can be sorted by the id or any /data/{index}/attributes key. | JSON pointer | Direction | sort | |--------------|-----------|------------| | /data/{index}/attributes/arrival_time | ascending | arrival_time | | /data/{index}/attributes/arrival_time | descending | -arrival_time | | /data/{index}/attributes/departure_time | ascending | departure_time | | /data/{index}/attributes/departure_time | descending | -departure_time | | /data/{index}/attributes/direction_id | ascending | direction_id | | /data/{index}/attributes/direction_id | descending | -direction_id | | /data/{index}/attributes/schedule_relationship | ascending | schedule_relationship | | /data/{index}/attributes/schedule_relationship | descending | -schedule_relationship | | /data/{index}/attributes/status | ascending | status | | /data/{index}/attributes/status | descending | -status | | /data/{index}/attributes/stop_sequence | ascending | stop_sequence | | /data/{index}/attributes/stop_sequence | descending | -stop_sequence | | /data/{index}/attributes/arrival_time if present, otherwise /data/{index}/attributes/departure_time | ascending | time | | /data/{index}/attributes/arrival_time if present, otherwise /data/{index}/attributes/departure_time | descending | -time |

  • :"fields[prediction]" (String.t): Fields to include with the response. Multiple fields MUST be a comma-separated (U+002C COMMA, ",") list. Note that fields can also be selected for included data types: see the V3 API Best Practices for an example.
  • :include (String.t): Relationships to include. schedule stop route trip vehicle alerts The value of the include parameter MUST be a comma-separated (U+002C COMMA, ",") list of relationship paths. A relationship path is a dot-separated (U+002E FULL-STOP, ".") list of relationship names. JSONAPI "include" behavior ## Example https://api-v3.mbta.com/predictions?filter%5Bstop%5D=place-sstat&filter%5Bdirection_id%5D=0&include=stop returns predictions from South Station with direction_id=0, below is a truncated response with only relevant fields displayed: { "data": [ { "id": "prediction-CR-Weekday-Fall-18-743-South Station-02-1", "relationships": { "stop": { "data": { "id": "South Station-02", "type": "stop" } }, }, "type": "prediction" } ], "included": [ { "attributes": { "platform_code": "2", }, "id": "South Station-02", "type": "stop" } ], } Note the stop relationship; use it to cross-reference stop-id with the included stops to retrieve the platform_code for the given prediction. ## Note on trips A Vehicle's trip is what is currently being served. A Prediction also has a vehicle: this is the vehicle we predict will serve this trip/stop. Since we know vehicles make future trips, the trip the vehicle is currently servicing can be different from the trips we're making predictions for. For example: Vehicle 1234 is currently serving trip A The block is Trip A → Trip B → Trip C We'll be making predictions for the rest of trip A, as well as all the stops of trip B and trip C. The trip for the Vehicle is always A, and all of the Predictions will reference Vehicle 1234.
  • :"filter[latitude]" (String.t): Latitude/Longitude must be both present or both absent.
  • :"filter[longitude]" (String.t): Latitude/Longitude must be both present or both absent.
  • :"filter[radius]" (Date.t): Radius accepts a floating point number, and the default is 0.01. For example, if you query for: latitude: 42, longitude: -71, radius: 0.05 then you will filter between latitudes 41.95 and 42.05, and longitudes -70.95 and -71.05.
  • :"filter[direction_id]" (String.t): Filter by direction of travel along the route. Must be used in conjuction with filter[route] to apply. The meaning of direction_id varies based on the route. You can programmatically get the direction names from /routes /data/{index}/attributes/direction_names or /routes/{id} /data/attributes/direction_names.
  • :"filter[route_type]" (String.t): Filter by route_type: https://developers.google.com/transit/gtfs/reference/routes-file. Multiple route_type MUST be a comma-separated (U+002C COMMA, ",") list. Must be used in conjunction with another filter.
  • :"filter[stop]" (String.t): Filter by /data/{index}/relationships/stop/data/id. Multiple IDs MUST be a comma-separated (U+002C COMMA, ",") list. Parent station IDs are treated as though their child stops were also included.
  • :"filter[route]" (String.t): Filter by /data/{index}/relationships/route/data/id. Multiple IDs MUST be a comma-separated (U+002C COMMA, ",") list.
  • :"filter[trip]" (String.t): Filter by /data/{index}/relationships/trip/data/id. Multiple IDs MUST be a comma-separated (U+002C COMMA, ",") list.
  • :"filter[route_pattern]" (String.t): Filter by /included/{index}/relationships/route_pattern/data/id of a trip. Multiple route_pattern_id MUST be a comma-separated (U+002C COMMA, ",") list.

Returns

• {:ok, MBTA.Model.Predictions.t} on success
• {:error, Tesla.Env.t} on failure