Install instructions for bioRad

During the Aeroecology course we will be using the R-package bioRad, a collection of tools to analyze and visualize biological signals in weather radar data. During the course you will get a sneak preview of the very latest version of bioRad (0.7.0), which hasn’t been published on CRAN yet.

Please take time to install the software before your travel to Colorado

We strongly recommend upgrading to a recent version of R (latest version is 4.2.1, found at https://www.r-project.org/). We also recommend installing Rstudio as a graphical user interface.

Please follow below install instructions to install the bioRad package (Note there are different instructions for Windows and for Mac OSX / Linux).

Installation on Windows systems

Step 1: install vol2birdR.

First, we install the vol2birdR package (a dependency package for bioRad). This package will be on CRAN soon, but we will install a pre-compiled binary for now:

install.packages(c("Rcpp", "assertthat", "pkgbuild", "rlang"))
install.packages("https://s3.amazonaws.com/mistnet/windows/vol2birdR_0.1.1.9913.zip",repos=NULL)

Step 2: download mistnet libraries

Next, we install several deep-learning and helper libraries to enable Mistnet”

vol2birdR::install_mistnet()

Step 3: download mistnet model

Next, we download the mistnet model into vol2birdR:

vol2birdR::install_mistnet_model()

Step 4: install bioRad

Finally, we can install bioRad:

# Step 4a: enable both the CRAN and bioconductor repositories:
setRepositories(ind=1:2)
# Step 4b: install latest release of bioRad:
install.packages("bioRad")
# Step 4c: upgrade bioRad to the latest development version:
install.packages("https://s3.amazonaws.com/mistnet/windows/bioRad_0.7.0.9522.zip", repos=NULL)

You should now be able to load bioRad. The welcome message should mention version 0.7.0.9522, and MistNet installed.

# load bioRad:
library(bioRad)

# expected message:
#> Welcome to bioRad version 0.7.0.9522
#> using vol2birdR version 0.1.1.9913 (MistNet installed)

Installation on Mac OSX and Linux systems:

Step 1: install system libraries:

We first need to install several system libraries that are used by vol2birdR package, which will be installed as part of bioRad.

Step 1 for Mac OSX:

First install Homebrew, a package manager for Mac OSX. It helps you installing missing libraries. After installing Homebrew, open a terminal (you can find it by searching for terminal) and type:

brew install hdf5 proj gsl

Step 1 for Linux:

On Debian-based systems (including Ubuntu), run:

# libraries required for vol2birdR:
sudo apt-get install libhdf5-dev libproj-dev gsl-bin libgsl-dev
# libraries required for other dependency packages:
sudo apt-get install libcurl4-openssl-dev libssl-dev libgdal-dev

On other Linux systems supporting yum and RPMs, run:

sudo yum install hdf5-devel proj-devel gsl gsl-devel

Step 2: pre-install some packages:

Next, open R using Rstudio. As of July 2022, a few packages require manual installation (for reasons unrelated to bioRad). In the console window execute the following:

# Step 2a: install package devtools (if you don't have it already)
install.packages("devtools")
# Step 2b: (Mac OSX only): install rgdal and terra
install.packages(c("rgdal","terra"), type="binary")
# Step 2c: upgrade ggmap to latest development version:
devtools::install_github("dkahle/ggmap")

Step 3: install bioRad

You can now open R using Rstudio and install bioRad as follows:

# Step 3a: enable both the CRAN and bioconductor repositories:
setRepositories(ind=1:2)
# Step 3b: install the latest development release of bioRad:
devtools::install_github("adokter/bioRad")

Step 4: download mistnet libraries

The bioRad installation of the previous step should have automatically installed the vol2birdR package as well, which is an important dependency package for bioRad. Next, we need to run a few vol2birdR commands to enable Mistnet (expect a ~ 150 Mb download)

vol2birdR::install_mistnet()

Step 5: download mistnet model

Next, we download the mistnet model into vol2birdR (expect >500 Mb download, this step can take a while when you have a slow internet connection)

vol2birdR::install_mistnet_model()

Step 6: load bioRad

You should now be able to load bioRad. The welcome message should mention version 0.7.0.9522, and MistNet installed.

# load bioRad:
library(bioRad)

# expected message:
#> Welcome to bioRad version 0.7.0.9522
#> using vol2birdR version 0.1.1.9913 (MistNet installed)