gce_vm_template.Rd
This lets you specify templates for the VM you want to launch It passes the template on to gce_vm_container
gce_vm_template(template = c("rstudio", "shiny", "opencpu", "r-base", "dynamic", "rstudio-gpu", "rstudio-shiny"), username = NULL, password = NULL, dynamic_image = NULL, image_family = "cos-stable", wait = TRUE, ...)
template | The template available |
---|---|
username | username if needed (RStudio) |
password | password if needed (RStudio) |
dynamic_image | Supply an alternative to the default Docker image for the template |
image_family | An image-family. It must come from the |
wait | Whether to wait for the VM to launch before returning. Default |
... | Arguments passed on to
|
The VM object, or the VM startup operation if wait=FALSE
Templates available are:
rstudio An RStudio server docker image with tidyverse and devtools
rstudio-gpu An RStudio server with popular R machine learning libraries and GPU driver. Will launch a GPU enabled VM.
rstudio-shiny An RStudio server with Shiny also installed, proxied to /shiny
shiny A Shiny docker image
opencpu An OpenCPU docker image
r_base Latest version of R stable
dynamic Supply your own docker image within dynamic_image
For dynamic
templates you will need to launch the docker image with any ports you want opened,
other settings etc. via docker_run.
Use dynamic_image
to override the default rocker images e.g. rocker/shiny
for shiny, etc.
if (FALSE) { library(googleComputeEngineR) ## make instance using R-base vm <- gce_vm_template("r-base", predefined_type = "f1-micro", name = "rbase") ## run an R function on the instance within the R-base docker image docker_run(vm, "rocker/r-base", c("Rscript", "-e", "1+1"), user = "mark") #> [1] 2 }