The config package makes it easy to manage environment specific
configuration values. For example, you might want to use distinct values
for development, testing, and production environments.
You can install the config package from CRAN by using:
install.packages("config")To use config, create a file config.yml with default as well as
other arbitrary configurations. For example:
config.yml
default:
trials: 5
dataset: "data-sampled.csv"
production:
trials: 30
dataset: "data.csv"To read configuration values you call the config::get function, which
returns a list containing all of the values for the currently active
configuration:
R
config <- config::get()
config$trials
#> [1] 5
config$dataset
#> [1] "data-sampled.csv"You can also read a single value from the configuration as follows:
R
config::get("trials")
#> [1] 5
config::get("dataset")
#> [1] "data-sampled.csv"The config::get() function takes an optional config argument which
determines which configuration to read values from (the “default”
configuration is used if none is specified).
See the package vignettes at https://rstudio.github.io/config/ for more examples and instructions for advanced usage, including:
- Inheritance and R expressions
- Using
configon Posit Connect
We strongly recommend you use config::get() rather than attaching the
package using library(config).
In fact, we strongly recommend you never use library(config).
The underlying reason is that the get() and merge() functions in
{config} will mask these functions with the same names in base R.