All repositories are available on GitHub under https://github.com/federated-partea

• MacOS Monterey 12.2.1 and Ubuntu 20.04
• The software was tested on both OS

1. Download the client software.
   1. macOS
   2. linux
   3. windows
2. Unzip the archive.
3. If you do not have Google Chrome installed already, please Download and install it.
4. Run the executable file
   1. macOS: Double-click on partea.app to start the client. The first time, you might need to run it with control + click.
   2. Ubuntu: Run the executable in your terminal with ./partea
5. The Partea Server and Partea Webapp are deployed at the servers of the Technical University Munich and do not need to be installed locally.

A getting started tutorial is available on the Partea website.

1. Sign up for an account.
2. [Sign in] with the account, if not done automatically after sign up.
3. Create a new project by clicking on the Projects tab, enter a project name and click Create.
4. Keep the default parameters to execute a non-parametric time-to-event analysis for survival function and cumulative hazard rate estimation. Initialize the project by clikcing the corresponding button.
5. Scroll down to the Collaborators section and press Create token. This is how you can invite participants to your study. This token can now be shared with a participant.

1. Now, we assume that we were a collaborator of the study. After receiving the token, we start our Partea client executable and sign in using our account credentials and the corresponding token.
2. Now we describe our dataset, which is in this case:
   1. Duration Column: week
   2. Event Column: arrest
   3. Seperator: comma
3. After that, we choose the dataset from our directory. You can download the sample data archive here. For simplicity, we will run the analysis here with only one collaborator, so we choose the dataset rossi_complete.csv, after unzipping the sample data archive.
4. Click "Run" to start the analysis. Now you wait, until the study coordinator starts the study. The client application should indicate this with the Message: "Waiting for other participants and project start."

1. As in this example we are the study coordinator, we go the Partea Webapp again, click on Projects and select our project.
2. After scrolling to the Collaborators section, we see that our client is running, and we can start the analysis by clicking Run Project.
3. After around 5-10 seconds, the analysis with only one client should be finished. The client application can be closed now, and the results are visible on the Webapp.
4. You can see all results in the webapp and compare the Survival Curve and Table with the expected results we attached here.

The Demo section basically explains already how to use Partea.

1. In the case of multiple participants, the coordinator simply creates more tokens and shares them with the corresponding participant.
2. Each participant then executes the client software locally, and press run.
3. After each participant has run the local computation, the study coordinator can run the project.