在RStudio中使用HERE API获取加拿大路线时单条结果全NA求助

在RStudio中调用HERE API计算加拿大境内驾车路线，代码如下：

# Packages
library("sf")
library("tidyverse")
library("hereR")
library("stringi") 
library("dplyr")

# Load data
distancePairs <- structure(list(Origin = c("Saskatoon", "St. John's"), Lat1 = c(52.1583630506965, 
47.5535718086621), Lon1 = c(-106.671978365698, -52.7486803304426
), Destination = c("St. John's", "Calgary"), Lat2 = c(47.5535718086621, 
51.0452926068494), Lon2 = c(-52.7486803304426, -114.05905562721
), id = c(17, 18), straight.distance.meters = structure(c(3808910.50512732, 
4332741.9186734), units = structure(list(numerator = "m", denominator = character(0)), class = "symbolic_units"), class = "units")), row.names = c(NA, 
-2L), class = c("tbl_df", "tbl", "data.frame"))

# Set API key
set_key(api_key = "YOUR_FREEMIUN_KEY") 

# Create SFDFs  (points with lat-lon pairs):
dist.pairs.input.1 <- dist.pairs.query %>%  
  st_as_sf(coords = c("Lon1","Lat1"), 
           crs = "WGS84", 
           remove = F)

dist.pairs.input.2 <- dist.pairs.query %>% 
  st_as_sf(coords = c("Lon2", "Lat2"), 
           crs = "WGS84", 
           remove = F)

#Calculate drive time&distance:
query_result <- route(origin = dist.pairs.input.1, 
                      destination = dist.pairs.input.2,
                      traffic = T)

result_summary <- query_result %>%
  group_by(id) %>% 
  summarise(new_shp = st_combine(.),
            departure = first(departure),
            arrival = last(arrival),
            type = mode(type),
            mode = mode(mode),
            distance_km = sum(distance)/1000,   
            duration_min = sum(duration)/60)   

distancePairs$id <- 1:2
result_output<- merge(result_summary,
                      dist.pairs.query,
                      by = "id",
                      all = TRUE)

运行后出现异常：第一行（萨斯卡通→圣约翰斯）返回完整的路线几何（MULTILINESTRINGZ）、距离和时长信息，但第二行（圣约翰斯→卡尔加里）所有字段为NA，几何对象显示为GEOMETRYCOLLECTION Z EMPTY。

• 变量名引用错误：代码中加载的数据集是distancePairs，但后续创建SF对象和合并结果时误用了未定义的dist.pairs.query，导致第二组坐标未正确传入API。
• 跨大陆路线的渡轮限制：圣约翰斯位于纽芬兰岛，到加拿大本土需乘坐渡轮，HERE API默认驾车路线不包含渡轮，无法直接计算跨岛的完整驾车路线。
• 长路线API限制：跨大陆超长路线可能触发API单次请求的计算阈值，或因实时流量数据覆盖不全导致请求失败。
• 批量请求容错机制：hereR的route()函数批量处理时，单个请求失败会返回NA而非中断，导致第二行结果缺失。

1. 修正变量名错误 #

统一使用加载的数据集名称distancePairs，替换所有误用的dist.pairs.query：

# 修正SF对象创建代码
dist.pairs.input.1 <- distancePairs %>%  
  st_as_sf(coords = c("Lon1","Lat1"), 
           crs = "WGS84", 
           remove = F)

dist.pairs.input.2 <- distancePairs %>% 
  st_as_sf(coords = c("Lon2", "Lat2"), 
           crs = "WGS84", 
           remove = F)

# 修正结果合并代码
result_output<- merge(result_summary,
                      distancePairs,
                      by = "id",
                      all = TRUE)

2. 启用渡轮支持并调整请求参数 #

显式允许渡轮路线，同时关闭实时流量（长路线流量数据覆盖有限）：

query_result <- route(origin = dist.pairs.input.1, 
                      destination = dist.pairs.input.2,
                      traffic = FALSE,
                      transport_mode = "car",
                      routing_mode = "fast",
                      allow_ferries = TRUE)

3. 拆分长路线并逐对处理 #

若跨大陆路线仍无法计算，将其拆分为两段（如圣约翰斯→哈利法克斯、哈利法克斯→卡尔加里），逐对请求后合并结果：

# 逐对处理路线请求，添加错误捕获
route_results <- map2(dist.pairs.input.1$geometry, dist.pairs.input.2$geometry, function(orig, dest) {
  tryCatch({
    route(origin = orig, destination = dest, traffic = FALSE, allow_ferries = TRUE)
  }, error = function(e) {
    message(paste("路线计算失败：", dist.pairs.input.1$Origin[which(dist.pairs.input.1$geometry == orig)], "→", dist.pairs.input.2$Destination[which(dist.pairs.input.2$geometry == dest)]))
    return(NULL)
  })
})

# 合并有效结果
query_result <- bind_rows(route_results)

4. 检查API配额与权限 #

确认HERE API密钥拥有路线计算服务权限，且剩余配额充足——超长路线计算会消耗更多配额，若配额不足会导致请求被限制。

内容的提问来源于stack exchange，提问作者MelaniaCB