使用 R 创建 GTFS 实时（车辆位置）

Question

我致力于将数据帧实时转换为 GTFS，并且在车辆位置部分苦苦挣扎。

我的数据看起来像这样（存储在一个名为“车辆”的数据框中）：

## Input data looks that way, one line per on-going vehicle
vehicle_id      trip_id   lat    lon bear speed           stop_time
52108 4.264930e+05 45.40 -71.92    1     9 2017-05-02 15:19:05
60105 4.273610e+05 45.40 -71.90  246     6 2017-05-02 15:18:59
59104 4.270150e+05 45.40 -71.87   81     7 2017-05-02 15:18:54

我的代码的详细信息是：

library(dplyr)
library(XML)
library(stringr)
library(RProtoBuf)
library(RODBC)

## Read the google gtfs proto file
readProtoFiles("gtfs-realtime.proto")

## List of current vehicles
current_vehicles <- unique(vehicle$vehicle_id)

## Create an empty list, 1 entry for each vehicle
protobuf_list <- vector(mode = "list", length = length(current_vehicles))

## Loop over all current vehicles
for(i in 1:length(current_vehicles)) {
  ## protobuf object 
  vehicle_position_update <- new(transit_realtime.VehiclePosition,
                                   vehicle = vehicle$vehicle_id[i],
                                   stop_id = vehicle$stop_id[i],
                                   trip = vehicle$trip_id[i],
                                   latitude = vehicle$lat[i],
                                   longitude = vehicle$lon[i],
                                   bearing = vehicle$bear[i],
                                   speed = vehicle$speed[i])

        ## protobuf feed entity
  e <- new(transit_realtime.FeedEntity,
           id = as.character(vehicle$vehicle_id[i]),
           vehicle = new(transit_realtime.VehiclePosition,
                             trip = new(transit_realtime.VehicleDescriptor,
                                        id = vehicle$vehicle_id[i]),
                         VehiclePosition = vehicle_position_update))

  ## Fill the list
  protobuf_list[[i]] <- e
}# Loop over vehicles

## GTFS header
header_object <- new(transit_realtime.FeedHeader,
                     gtfs_realtime_version = "1.0",
                     incrementality = "FULL_DATASET",
                     timestamp = as.numeric(as.POSIXlt(Sys.time())))

## Build the full GTFS
m <- new(transit_realtime.FeedMessage,
         header = header_object,
         entity = protobuf_list) # use entity_list

## Write the GTFS
writeLines(as.character(m))

## Turn it into binary
serialize(m, "vehiclePositions.pb")

在创建 protobuffer 对象 vehicle_position_update 时，它​​会崩溃并显示以下消息：

  type mismatch, expecting a 'Message' object

我浏览了 gtfs-realtime.proto，我对要包含的不同消息的理解似乎很好（嗯，显然不是......）。

有人知道为什么无法创建这个 protobuffer 文件吗？

为明确的解决方案添加：

我的问题是我没有完全遵循不同消息的 gtfs 原型描述。一旦这一点得到纠正，车辆上的循环就变成了：

## Loop over all current vehicles
for(i in 1:length(current_vehicles)) {
    ## protobuf object
     vehicle_position_update <- new(transit_realtime.Position,
                                   latitude = vehicle$lat[i],
                                   longitude = vehicle$lon[i],
                                   bearing = vehicle$bear[i],
                                   speed = vehicle$speed[i])
        ## protobuf feed entity
  e <- new(transit_realtime.FeedEntity,
           id = as.character(vehicle$vehicle_id[i]),
           vehicle = new(transit_realtime.VehiclePosition,
                             trip = new(transit_realtime.TripDescriptor,
                                        trip_id = vehicle$trip_id[i],
                                        route_id = vehicle$route_id[i]),
                         stop_id = vehicle$stop_id[i],
                         position = vehicle_position_update))
  ## Fill the list
  protobuf_list[[i]] <- e
}# Loop over vehicles

它可以工作

Answer 1

消息定义告诉你它需要哪些字段，例如

writeLines(as.character(RProtoBuf::fileDescriptor(transit_realtime.FeedMessage)))

message FeedMessage {
  required .transit_realtime.FeedHeader header = 1;
  repeated .transit_realtime.FeedEntity entity = 2;
  extensions 1000 to 1999;
}

message FeedHeader {
  enum Incrementality {
    FULL_DATASET = 0;
    DIFFERENTIAL = 1;
  }
  required string gtfs_realtime_version = 1;
  optional .transit_realtime.FeedHeader.Incrementality incrementality = 2 [default = FULL_DATASET];
  optional uint64 timestamp = 3;
  extensions 1000 to 1999;
}

message FeedEntity {
  required string id = 1;
  optional bool is_deleted = 2 [default = false];
  optional .transit_realtime.TripUpdate trip_update = 3;
  optional .transit_realtime.VehiclePosition vehicle = 4;
  optional .transit_realtime.Alert alert = 5;
  extensions 1000 to 1999;
}
... etc

然后，如果您查看Position 消息，您会看到字段

message Position {
  required float latitude = 1;
  required float longitude = 2;
  optional float bearing = 3;
  optional double odometer = 4;
  optional float speed = 5;
  extensions 1000 to 1999;
}

所以你使用这些值定义Position，例如

RProtoBuf::new(transit_realtime.Position, latitude = 0, longitude = 0)  

VehiclePosition 消息是

message VehiclePosition {
  enum VehicleStopStatus {
    INCOMING_AT = 0;
    STOPPED_AT = 1;
    IN_TRANSIT_TO = 2;
  }
  enum CongestionLevel {
    UNKNOWN_CONGESTION_LEVEL = 0;
    RUNNING_SMOOTHLY = 1;
    STOP_AND_GO = 2;
    CONGESTION = 3;
    SEVERE_CONGESTION = 4;
  }
  enum OccupancyStatus {
    EMPTY = 0;
    MANY_SEATS_AVAILABLE = 1;
    FEW_SEATS_AVAILABLE = 2;
    STANDING_ROOM_ONLY = 3;
    CRUSHED_STANDING_ROOM_ONLY = 4;
    FULL = 5;
    NOT_ACCEPTING_PASSENGERS = 6;
  }
  optional .transit_realtime.TripDescriptor trip = 1;
  optional .transit_realtime.VehicleDescriptor vehicle = 8;
  optional .transit_realtime.Position position = 2;
  optional uint32 current_stop_sequence = 3;
  optional string stop_id = 7;
  optional .transit_realtime.VehiclePosition.VehicleStopStatus current_status = 4 [default = IN_TRANSIT_TO];
  optional uint64 timestamp = 5;
  optional .transit_realtime.VehiclePosition.CongestionLevel congestion_level = 6;
  optional .transit_realtime.VehiclePosition.OccupancyStatus occupancy_status = 9;
  extensions 1000 to 1999;
}

所以消息会是这样的

RProtoBuf::new(transit_realtime.VehiclePosition, 
                             current_status = 1,
                             congestion_level = 0,
                             stop_id = "7", 
                             current_stop_sequence = 1)