How to backup data with git

git is a distributed version control system that allows us to keep tight control of our data. Once the data are tracked with git, we will see any changes to the data and be able to track any unintended overwrites or deletion of participant data. This guide is using github desktop to keep the data under control and give us peace of mind while analyzing data.

Step 0: collect the data

The steps to back up the data will be completed after the session with the participant is over.

Step 1: Open github desktop

Open github desktop, the desktop icon should be on the main screen of your computer.

Step 2: Fetch changes from github

The correct repository should already be open (however, there may be multiple repositories to switch between). We want to make sure our local copy of the data is the most up to date version, so we will ask the server (github) to "fetch" any changes. If there are changes that were pulled from the server, to apply them to the local repository, you will have to click that button again, the button should change from saying fetch to saying pull.

Step 3: Add/Commit changed files

If files have been changed in the repository, github desktop should already see those changed files and have them checked. We want to keep track of those changes so we will write a commit message so we can make a summary of what files are being added/modified. Most commonly, your commit message will be: add sub-<label> data, where <label> is the subject study identifier. In the description box, you can provide a more detailed explanation of the changes being made, in case additional files were modified unexpectedly. Or in case something went wrong or strange with the data collection.

Step 4: Make the commit

Once the title and description of the commit is set, select the commit to master button.

Step 5: push changes to origin

Finally we will push the changes to github, backing up our data and making it easily accessible to other lab members to analyze the data.

Step 6: repeat steps 2-5

Repeat this process for each "repository" (i.e. generally each task), Until all the tasks have been uploaded to github.

This process is important to maintain validity of the data and makes the steps from data collection to pretty graphs to impactful results much faster, so thank you for all your help!

For questions please contact: [email protected] via this email, in slack, or in person.