Install Python dependencies for greta

This is a helper function to install Python dependencies needed. This includes Tensorflow version 1.14.0, Tensorflow Probability 0.7.0, and numpy version 1.16.4. These Python modules will be installed into a virtual or conda environment, named "greta-env". Note that "virtualenv" is not available on Windows.

install_greta_deps(
  method = c("auto", "virtualenv", "conda"),
  conda = "auto",
  timeout = 5,
  ...
)

Arguments

method	Installation method ("virtualenv" or "conda")
conda	The path to a conda executable. Use "auto" to allow reticulate to automatically find an appropriate conda binary. See Finding Conda for more details.
timeout	maximum time in minutes until the installation for each installation component times out and exits. Default is 5 minutes per installation component.
...	Optional arguments, reserved for future expansion.

Note

This will automatically install Miniconda (a minimal version of the Anaconda scientific software management system), create a 'conda' environment for greta named 'greta-env' with required python and python package versions, and forcibly switch over to using that conda environment.

If you don't want to use conda or the "greta-env" conda environment, you can install these specific versions of tensorflow (version 1.14.0), and tensorflow-probability (version 0.7.0), and ensure that the python environment that is initialised in this R session has these versions installed. This is now always straightforward, so we recommend installing the python packages using install_greta_deps() for most users.

Examples

# NOT RUN {
install_greta_deps()
# }