使用amt包嵌套GPS数据时遇atan2参数缺失错误求助

背景 #

使用2019-2022年哈德逊湾西部北极熊GPS位置数据集，按照amt包官方vignette流程处理GPS追踪数据时，示例数据可正常运行，但自有数据在嵌套步骤后调用amt函数出现错误。非嵌套子集可正常运行，怀疑amt函数列识别存在问题。

运行代码 #

# Load libraries
library(sf)
library(lubridate)
library(raster)
library(move)
library(amt) 
library(tibble)
library(dplyr)


# Load data
polarBear_GPS <- polarBear_GPSLocation_westHudsonBay_2019_2022
View(polarBear_GPS)

ind<-complete.cases(polarBear_GPS[,c("Datetime", "Longitude", "Latitude")])
polarBear_GPS<-polarBear_GPS[ind==TRUE,]

# Remove NA values from Datetime
polarBear_GPS <- polarBear_GPS[!is.na(polarBear_GPS$Datetime), ]

# Check for NA values again, should return 0
sum(is.na(polarBear_GPS$Datetime))  # Should return 0

# Make timestamp a date/time variable
polarBear_GPS$Datetime <- as.POSIXct(polarBear_GPS$Datetime, format="%m/%d/%Y %H:%M:%S", tz="UTC")


# Create the track with WGS 84 CRS
trk1 <- make_track(
  polarBear_GPS,
  .x = Longitude,
  .y = Latitude,
  .t = Datetime,
  id = Bear,
  crs = st_crs(4326)
)

# Calculate day/night
trk1 <- trk1 %>% time_of_day()

# Transform coordinates (此处CRS未变化，可省略)
trk1 <- transform_coords(trk1, crs_from = 4326, crs_to =4326 )

# Nest data by individual
trk2 <- trk1 %>% nest(data = -id)

# Attempt to calculate movement metrics (报错代码)
trk3 = trk2 %>% mutate(dir_abs = purrr:::map(data, direction_abs,full_circle=TRUE, zero="N"), 
                       dir_rel = purrr:::map(data, direction_rel), 
                       sl = purrr:::map(data, step_lengths),
                       nsd_=purrr:::map(data, nsd))

数据结构 #

str(trk1)
Classes ‘track_xyt’, ‘track_xy’, ‘spec_tbl_df’, ‘tbl_df’, ‘tbl’ and 'data.frame':   118858 obs. of  5 variables:
 $ x_  : num  -93.2 -93.2 -93.2 -93.2 -93.2 ...
 $ y_  : num  58.8 58.8 58.8 58.8 58.8 ...
 $ t_  : POSIXct, format: "2019-08-26 20:20:11" "2019-08-26 20:25:10" "2019-08-26 20:30:09" ...
 $ id  : chr  "X17517" "X17517" "X17517" "X17517" ...
 $ tod_: Factor w/ 2 levels "day","night": 1 1 1 1 1 1 1 1 1 1 ...
 - attr(*, "spec")=List of 3
  ..$ cols   :List of 4
  .. ..$ Bear     : list()
  .. .. ..- attr(*, "class")= chr [1:2] "collector_character" "collector"
  .. ..$ Datetime : list()
  .. .. ..- attr(*, "class")= chr [1:2] "collector_character" "collector"
  .. ..$ Latitude : list()
  .. .. ..- attr(*, "class")= chr [1:2] "collector_double" "collector"
  .. ..$ Longitude: list()
  .. .. ..- attr(*, "class")= chr [1:2] "collector_double" "collector"
  ..$ default: list()
  .. ..- attr(*, "class")= chr [1:2] "collector_guess" "collector"
  ..$ delim  : chr ","
  ..- attr(*, "class")= chr "col_spec"
 - attr(*, "problems")=<externalptr> 
 - attr(*, "crs_")= num 4326


str(trk2)
nstd_trc [20 × 2] (S3: nested_track/tbl_df/tbl/data.frame)
 $ id  : chr [1:20] "X17517" "X19911" "X33824" "X33939" ...
 $ data:List of 20
  ..$ : trck_xyt [5,435 × 4] (S3: track_xyt/track_xy/spec_tbl_df/tbl_df/tbl/data.frame)
  .. ..$ x_  : num [1:5435] -93.2 -93.2 -93.2 -93.2 -93.2 ...
  .. ..$ y_  : num [1:5435] 58.8 58.8 58.8 58.8 58.8 ...
  .. ..$ t_  : POSIXct[1:5435], format: "2019-08-26 20:20:11" "2019-08-26 20:25:10" "2019-08-26 20:30:09" "2019-08-26 20:35:09" ...
  .. ..$ tod_: Factor w/ 2 levels "day","night": 1 1 1 1 1 1 1 1 1 1 ...
  .. ..- attr(*, "spec")=List of 3
  .. .. ..$ cols   :List of 4
  .. .. .. ..$ Bear     : list()
  .. .. .. .. ..- attr(*, "class")= chr [1:2] "collector_character" "collector"
  .. .. .. ..$ Datetime : list()
  .. .. .. .. ..- attr(*, "class")= chr [1:2] "collector_character" "collector"
  .. .. .. ..$ Latitude : list()
  .. .. .. .. ..- attr(*, "class")= chr [1:2] "collector_double" "collector"
  .. .. .. ..$ Longitude: list()
  .. .. .. .. ..- attr(*, "class")= chr [1:2] "collector_double" "collector"
  .. .. ..$ default: list()
  .. .. .. ..- attr(*, "class")= chr [1:2] "collector_guess" "collector"
  .. .. ..$ delim  : chr ","
  .. .. ..- attr(*, "class")= chr "col_spec"
  .. ..- attr(*, "problems")=<externalptr> 
  .. ..- attr(*, "crs_")= num 4326

错误信息 #

Error in mutate():
ℹ In argument: dir_abs = purrr:::map(data, direction_abs, full_circle = TRUE, zero = "N").
Caused by error in purrr:::map():
ℹ In index: 1.
Caused by error in h():
! error in evaluating the argument 'y' in selecting a method for function 'atan2': argument "y" is missing, with no default

中文翻译

在mutate()中出错：
ℹ 出错参数：dir_abs = purrr:::map(data, direction_abs, full_circle = TRUE, zero = "N")。
由purrr:::map()引发错误：
ℹ 出错索引：1。
由h()引发错误：
! 为函数atan2选择方法时，计算参数'y'出错：参数"y"缺失，无默认值

解决方案 #

问题出在两个点：

1. 使用了purrr:::map（三个冒号）调用purrr的内部未导出函数，这是非标准用法，容易引发参数传递问题；
2. 直接将direction_abs等amt函数传入map时，参数传递逻辑不符合amt函数的要求，需要显式将嵌套的track对象传递给函数。

修改后的代码如下：

# 使用标准的purrr::map，并通过公式语法显式传递track对象
trk3 <- trk2 %>% 
  mutate(
    dir_abs = purrr::map(data, ~direction_abs(.x, full_circle = TRUE, zero = "N")),
    dir_rel = purrr::map(data, ~direction_rel(.x)),
    sl = purrr::map(data, ~step_lengths(.x)),
    nsd_ = purrr::map(data, ~nsd(.x))
  )

另外，也可以使用amt包提供的更简洁方式，无需手动嵌套后调用purrr：

# 直接按id分组计算所有指标
trk_metrics <- trk1 %>%
  group_by(id) %>%
  mutate(
    dir_abs = direction_abs(full_circle = TRUE, zero = "N"),
    dir_rel = direction_rel(),
    sl = step_lengths(),
    nsd_ = nsd()
  ) %>%
  ungroup()

内容的提问来源于stack exchange，提问作者Ellery Vincent