wilddogr

wilddogr is an R-package that allows you to automatically download and clean GPS data of wild dogs stored on Dropbox. For this, you’ll need to have access to the “KML-Dispersers” folder.

Installation

To install wilddogr you need to have the package devtools installed. You can then install wilddogr using:

devtools::install_github("DavidDHofmann/wilddogr")

Workflow

The wilddogr package follows a two-step process using the functions, dog_files() and dog_download().

1. dog_files() This function is used to retrieve information on all files containing GPS data in the KML-Dispersers folder on Dropbox. You can use the returned dataframe to subset to the files you’re interested in. The dataframe then serves as input to the dog_download() function.

2. dog_download() Once you have selected the files you’re interested in, download them using dog_download(). The function allows you to specify if you want to download the raw data (clean = F) or if you’d like to have the data pre-processed (clean = T). By default, the function directly returns the filepaths to the downloaded files so that you can easily load it into your r-session.

In addition, there’s a helper function dog_download_all() that serves to simply download all GPS data that’s available on Dropbox. For further details and options, check the help files for the different functions.

Finally, the function resampleFixes() can be used to coarsen the regularize the temporal resolution of the GPS data to a specified value (e.g. 4-hours or 24-hours).

Example

Here is a little example of the above outlined workflow

# Load required packages
library(wilddogr)
library(ggpubr)

# Identify files on dropbox (including the rvc data)
files <- dog_files(rvc = T)

# Let's take a look at the files we could download
head(files)

# Subset to the files of interest
todownload <- subset(files, DogName %in% c("Abel", "Tangiers"))

# Download the data
downloaded <- dog_download(
    x         = todownload
  , clean     = T
  , overwrite = T
  , outdir    = getwd()
  , printpath = T
)

# Let's take a look at the data
dat <- read_csv(downloaded)
head(dat)

# We can also coarsen it to 24 hours
dat_res <- resampleFixes(dat, hours = 24, start = 7, tol = 0.5)

# Plot the original and resampled data
p1 <- ggplot(dat, aes(x = x, y = y, col = as.factor(DogName))) +
  geom_path() +
  geom_point(size = 0.8) +
  coord_equal() +
  theme_minimal() +
  scale_color_manual(values = c("orange", "cornflowerblue"), name = "DogName") +
  ggtitle("Original")
p2 <- ggplot(dat_res, aes(x = x, y = y, col = as.factor(DogName))) +
  geom_path() +
  geom_point(size = 0.8) +
  coord_equal() +
  theme_minimal() +
  scale_color_manual(values = c("orange", "cornflowerblue"), name = "DogName") +
  ggtitle("Resampled (24 Hours)")
ggarrange(p1, p2, common.legend = T)