Download and format phenocam sites in phenor

format_phenor_data.R

# required libraries
library(phenor)

# where to store the data
out_dir <- "data-raw/"

# sites to process, stored in a csv file
# with columns site, veg_type, roi (check for extra spaces
# which can mess with name formatting)
sites <- read.table("data/sites.csv", sep = ",", header = TRUE)

# loop over all sites and download and unzip
# nightly builds (intermediate product releases
# from the official phenocam US website
# the same can be done using the phenocamr package
# but might take longer depending on the number of sites
# and required control over processing parameters)
apply(sites, 1, function(site){
  
  provisional_data <- sprintf("%s_%s_%s_provisional_data.zip",
                              site['site'],
                              site['veg_type'],
                              site['roi'])
  
  url <- file.path("https://phenocam.sr.unh.edu/data/archive",
                   site['site'], "ROI", provisional_data)
  tmp_file <- file.path(tempdir(),"tmp.zip")
  dates <- sprintf(
    "provisional_data/data_record_5/%s_%s_%s_3day_transition_dates.csv",
    site['site'],
    site['veg_type'],
    site['roi'])
  
  error <- try(download.file(url = url,
                             destfile = tmp_file))
  
  if(!inherits(error,"try-error")){
    unzip(tmp_file,
          files = dates, 
          junkpaths = TRUE,
          exdir = "data/"
          )
  }
})

# format all files into one phenor dataset for training models
phenocam_transition_dates <- pr_fm_phenocam(
  path = out_dir,
  direction = "rising",
  gcc_value = "gcc_90",
  threshold = 50,
  offset = 264,
  internal = TRUE
)

# save phenocam data
saveRDS(phenocam_transition_dates, "data/phenocam_data.rds")