Working locally

This section describes the general steps to setup this project on most UNIX systems.

1. Requirements

  • R scripts and R notebooks:

    • R 3.5.2

    • RStudio or RStudio server

  • IPython notebooks:

    • Python 3

    • Python libraries:

      renku-r-tools==0.0.3
slims-lisp==0.1.3

2. Get the R project “template”

Use git to clone this project where you need it.

$ git clone https://github.com/auwerxlab/survival_analysis.git

3. Update the README file

Now that the project is ready to start, update its README.rst file with your author’s information and a short description.

4. Install R packages

This R project uses packrat to manage R packages.

The R packages sources are provided along with this project. However, compiled R libraries are not.

Therefore, the R libraries first need to be compiled from the provided sources using the packrat::restore() command.

Run the following commands in the R console to enable packrat and install the required libraries:

> packrat::on()
> packrat::restore()

5. Keep track of your work

Danger

Understand what you are doing. Hosting your project on the wrong repository can expose sensitive information and data! 💥

It usually boils down to the following points:

  • Understand the project’s privacy requirments.

  • Know who has access to the git repository.

  • If you whitness a breach, immediatly inform the responsible persons and fix the breach (make sure to also delete all sensitive information from previous versions and logs).

Use a version control system like git to keep track of your work.

First, create a new git remote repository (ex: on GitHub).

Then, set its url in the project:

$ git remote set-url origin https://<your_new_remote_url>.git

When you are ready to save changes made to the project, take a snapshot of all its files:

$ git add -A

Then, commit the changes:

$ git commit -m "<your_short_description>"

And finally, push the committed changes to the remote git repository:

$ git push origin master

6. Keep up to date with the last version of the project “template”

To keep updated with the last version of the code, set an upstream remote repository and use the git rebase command.

Set the upstream remote repository:

$ git remote add upstream https://github.com/auwerxlab/survival_analysis.git

Whenever you need the latest updates from the upstream repository, run:

$ git fetch upstream
$ git rebase upstream/master

👍 That is it!