If you’re using renv with an R project that also depends on some Python packages (say, through the reticulate package), then you may find renv’s Python integration useful.
Activating Python integration
Python integration can be activated on a project-by-project basis.
Use renv::use_python() to tell renv to create and use a
project-local Python environment with your project. If the reticulate
package is installed and active, then renv will use the same version of
Python that reticulate normally would when generating the virtual
environment. Alternatively, you can set the
RETICULATE_PYTHON environment variable to instruct renv to
use a different version of Python.
If you’d rather tell renv to use an existing Python virtual
environment, you can do so by passing the path of that virtual
environment instead – use
renv::use_python(python = "/path/to/python") and renv will
record and use that Python interpreter with your project. This can also
be used with pre-existing virtual environments and Conda
environments.
Understanding Python integration
Once Python integration is active, renv will attempt to manage the
state of your Python virtual environment when snapshot() /
restore() is called. With this, projects that use renv and
Python can ensure that Python dependencies are tracked in addition to R
package dependencies. Note that future restores will require both
renv.lock (for R package dependencies) and
requirements.txt (for Python package dependencies).
Virtual environments
When using virtual environments, the following extensions are provided:
renv::snapshot()callspip freeze > requirements.txtto save the set of installed Python packages;renv::restore()callspip install -r requirements.txtto install the previously-recorded set of Python packages.
Conda environments
When using Conda environments, the following extensions are provided:
renv::snapshot()callsconda env export > environment.ymlto save the set of installed Python packages;renv::restore()callsconda env [create/update] --file environment.ymlto install the previously-recorded set of Python packages.