workspace

A package for allowing easy management of your project workspace.

Installation

You can install the package directly from GitHub:

devtools::install_github("rcannood/workspace")

You also need to set an environment variable MY_WORKSPACE_PATH. For unix-based systems, this can be done by placing the following content in ~/.Renviron:

MY_WORKSPACE_PATH=/home/user/Workspace

Restart R to ensure the environment variables are correctly set.

Explanation

You can create a new project as follows:

library(workspace)
workspace::initialise_project("some_project")

The following folders will be created:

• /home/user/Workspace/some_project/:
  • scripts/: Scripts to process raw data or derived data into results.
  • R/: Reusable functions which will be compiled as a package named some_project.
  • raw_data/: Any data you created or download that you certainly never want to delete
  • derived_data/: Data reproducible by scripts and raw data; you should be able to safely delete this folder.
  • results/: Any figures or data objects that you save.

Example

An R script located at /home/user/Workspace/some_project/scripts/script.R could contain the following:

library(tidyverse)
library(workspace)

# let workspace know what experiment you're working on
project("some_project", experiment_id = "my_first_experiment")

# save data at derived_data/my_first_experiment/data.rds
data <- tribble(
  ~x, ~y, 
  1,  2,
  3,  1, 
  7,  2
)
write_rds(data, derived_file("data.rds"))

# save a plot at results/my_first_experiment/myplot.pdf
g <- ggpot(data) + geom_point(aes(x, y))
ggsave(result_file("myplot.pdf"), g, width = 10, height = 10)

# read data from a different project, located at derived_data/some_other_project/data.rds
old_data <- read_rds(derived_file("data.rds", "some_other_project"))

# read this script... for whatever reason.
code_txt <- read_lines(script_file("script.R"))