In cboettig/movebank: Lightweight access to 'Movebank.org' data from animal GPS trackers

movebank

knitr::opts_chunk$set(message=FALSE)

Use

library("movebank")
library("tidyverse")
library("sf")

readRenviron("../../.Renviron")

For some endpoints we can successfully get the full table in a single call:

study <- movebank("study") ## temporal & geographic coverage, but no species coverage information
taxon <- movebank("taxon") ## Purely a taxonomic database, has no connection to study
tag_type <- movebank("tag_type") ## small table listing information the different types of animal tracker tags

The 46 datasets for which public data actually seems to be avaliable. (the study table suggests r sum(study$i_can_see_data =="true") datasets have i_can_see_data, but querying on those datasets finds otherwise. Maybe the others have some license condition to accept first? Still, notable inconsistencies between the use of this flag and the actual data download, see examples.md notes).

public <- read_csv("../../public_ids.csv")[[1]]

Okay, we're ready to download the main data files for these studies. The movement data itself is found in the event table. More information about the individuals involved in the study, including the species identification, is given separately in the individuals table.

## helper function to bind all results into single data frame.
movebank_df <- function(ids, entity){
  df <- 
    map(ids, 
      safely(function(i) 
        movebank(entity, study_id = i, guess_types=FALSE))) %>%
    map_df("result", .id = "study")

  ## create a column with study id as the name
  tmp <- data_frame(study_id = ids, study = as.character(1:length(ids)))
  select(left_join(df, tmp, "study"), -study)
}

individual <- movebank_df(public, "individual")
tag <- movebank_df(public, "tag")
event <- movebank_df(public, "event")

cboettig/movebank documentation built on May 24, 2019, 3:05 a.m.