knitr::opts_chunk$set( collapse = TRUE, comment = "#>", eval = TRUE )
prodigenr, or project directory generator, was designed to simplify the process of creating new scientific data analysis projects and to help make your workflow more reproducible and open from the beginning. While creating individual folders and files manually for new projects doesn't take too much time, over time and over many researchers, this can quickly add up to a lot of time! Plus, when a standard structure is followed it makes it easier to share code and establish reproducible practices earlier on in the project.
Starting a research project? Create a project directory like so:
library(prodigenr) # Create a temporary folder using the fs package new_project_path <- fs::path_temp("HeartDiseaseExercise") setup_project(new_project_path)
Or via RStudio's interface (with RStudio version >1.1):
The resulting file structure should look something like this:
prodigenr:::viz_project_tree(new_project_path)
README.md
files are contained within each project and in each folder that explains
a bit more about what each folder and file are used for, which is briefly described
here:
R/
: Should contain the R scripts and functions used for the analysis.doc/
: Should contain the files related to presenting the project's scientific
output. Already has the report/manuscript inside.data/
: If relevant, is where the processed (or simulated) data is kept that
is used for the project as well as the results of the project's analyses.data-raw/
: If relevant, is where the scripts that process the raw data into
the usable data are kept and, optionally where the raw data is also kept.DESCRIPTION
: Is a standard file that includes metadata about your project, in
a machine readable format, and that also stores a list of the R packages your project
depends on..Rproj
: Is a standard file used by RStudio to set some R Project specific
settings.To add a new document (e.g. slides, manuscript), run any of the create_*()
commands (e.g. create_slides()
) in the console while in RStudio in the
newly created project (via the .Rproj
file):
# you need to run these in the project's console create_slides()
# you need to run these in the project's console withr::with_dir( new = new_project_path, code = { create_slides() } )
Now two more files have been added to the doc/
folder. The resulting file
structure should look something like this:
withr::with_dir(fs::path_temp(), {fs::dir_tree(basename(new_project_path))})
At present, there are only two template files that you can view:
template_list
These template files are what an academic researcher likely typically encounters. However, if you have a suggestion or want to add a template, please create a Github issue or submit a Pull Request!
The end goal of each project is to be as self contained as possible. So that if you ever need to go back to the analysis, it is easy to re-run the code and get the results that you say you got. This is especially useful if others such as reviewers ask for something or want to confirm your results. See the manifesto for more details on the underlying philosophy behind this package.
There are several ways of handling a project. There a few packages that have
similar functionality as prodigenr
package structure:
ProjectTemplate
is well documented and still
actively developed. Only downside is that it is fairly complicated to use and
complex in the project workflow it creates.makeProject
is very simple and stripped down in what it creates and in it's use. Downside is
that it wasn't updated since 2012.devtools
(or
usethis
),
which is argued for in this
blog and
compared to ProjectTemplate
in this
blog).rrtools
is very similar to prodigenr,
except it focuses only on manuscripts. Is well thought out and the documentation
is well written.workflowr
is a workflow for creating
online, data science content.There is also a list of other similar projects on the rOpenSci GitHub repository. It's up to you to decide which style to use.
fs::dir_delete(new_project_path)
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.