CRAN version (cautious users)

The stable version of oce is provided on the R archives. It is installed from within R with


Pre-compiled development version

Occasionally, oce is compiled for MSWindows and for OS X. The results are stored at a Dropbox site. Visit that site and the appropriate subdirectory (osx or windows), and download both oce and ocedata. Then install as R normally installs packages, e.g. on an OS X machine, type the following in a terminal (console) window:

R CMD install oce_0.9-22.tgz 
R CMD install ocedata_0.1.4.tgz 

Note that if your system uncompresses the files, you will have to change the suffix .tgz to .tar to make the above-stated commands work.

Users of the RStudio IDE can install the package from the downloaded package files by using the menu item Install Packages under the Tools menu, and selecting Install from: Package archive file. A similar approach can be used with the GUI under the Packages & Data menu.

Source-code development version

Users who are set up to build packages (which requires a C compiler, etc; see note 1) may stay at the cutting edge by installing a development version. This may be done in several ways, of which the third is the best for most users.

  1. Download the tarball sources (oce and ocedata), expand the files, rename the resultant directories to oce and ocedata and then execute the OS commands stated below.

    R CMD build oce
    R CMD install oce_0.9-19.tar.gz # adjust version as needed
    R CMD build ocedata
    R CMD install ocedata_0.1.4.gz  # adjust version as needed

    A similar procedure works for the zip files for oce and ocedata.

  2. Clone the sources with Git

    git clone
    git clone

    and then build as described above.

  3. Build with devtools::install_github()

    install_github("dankelley/oce", ref="develop")
    install_github("dankelley/ocedata", ref="master")

The first method is included here for convenience of users who do not have git on their computers. The second method is the one chosen by most users who have programming experience. The third method is recommended for most users who simply want to build an up-to-date version of oce.

Note. Getting ready to compile (by installing a C compiler, etc.) is easy on linux, although the method depends on the system. On MSWindows, install Rtools from CRAN. On OSX, it can be fairly complicated, as explained in a wiki page.

This website is written in Jekyll, and the source is available on GitHub.