Accelerator Docs

Appearance

Getting Started with Accelerator ​

Welcome to the Accelerator Platform — an integrated system for running scientific computations, data pipelines, and managing reusable models.

This guide provides a high-level overview of how to get started, what the user experience looks like, and introduces the core concepts of the platform.

1️⃣ Sign In ​

Platform URL: https://accelerator.iiasa.ac.at

Before using Accelerator, you need to sign in.

  • IIASA users → can sign in with their IIASA account.
  • External users → can register and then sign in.
Accelerator Sign In Screen
Accelerator Sign In Screen

2️⃣ Terminal CLI Installation ​

Installation ​

No Python install required. Run the appropriate command for your system:

Linux / macOS:

bash
curl -fsSL https://raw.githubusercontent.com/iiasa/accli/master/scripts/install.sh | bash

Windows (PowerShell):

powershell
irm https://raw.githubusercontent.com/iiasa/accli/master/scripts/install.ps1 | iex

Option 2: Via PIP ​

bash
pip install accli --user

Authentication ​

To authenticate your terminal session with the Accelerator platform:

bash
accli login

Follow the on-screen instructions to complete the sign-in process in your browser.

3️⃣ Project Spaces ​

After signing in, you will land on the Projects List Page.

  • Here you will see a list of Project Space cards for spaces you are a member of.
  • If you do not see any project spaces:
    • You are not yet added to a space.
    • You can create a new Project Space (if permitted), or
    • Ask a Project Space admin to add you.

Once added to any Project Space, it will appear on your list.

Key features of Project Spaces:

  • Dedicated, isolated team workspace: Each project space acts as a secure sandbox, ensuring that datasets, models, and computational results remain isolated from other projects.
  • Scalable team membership: Add or remove members as your collaboration grows, with support for varying levels of access and control within the project environment.
  • Role-based resource access: Manage permissions at a granular level by assigning specific roles (viewer, editor, admin) to control how members interact with files, routines, and jobs.
Project Space List
Project List in Accelerator
Project Space Overview
Project Space Overview in Accelerator

4️⃣ Major Features inside a Project Space ​

Inside each Project Space, you will find a set of key features available via the main menu:

Templates ​

  • Contains a list of Dataset Templates.
  • Templates define validation rules for specific datasets.
  • Supports CSV Timeseries, Regional Timeseries, and Raster Timeseries.

Detailed documentation → Dataset Validation and Templates

Files ​

  • Holds all files and folders stored in the Project Space.
  • Supports uploading, downloading, and various file actions.
  • Files are stored close to the compute cluster, enabling efficient I/O during computations.

Detailed documentation → Files and Data Management

Jobs ​

  • Displays the status of all jobs run in the Project Space.
  • Includes jobs from individual routines and jobflows.
  • Users can view logs, monitor progress, and access job outputs.

Detailed documentation → Job Monitoring and Logs

Routines ​

  • Holds the definition of computational tasks (Routines).
  • Users can create, modify, and launch routines as jobs.
  • Routines can be used standalone or as building blocks for jobflows.

Detailed documentation → Working with Routines

Jobflows ​

  • Manage jobflows — graphs of routines.
  • Users can create, modify, and run jobflows.
  • Supports both simple and complex acyclic graphs.

Detailed documentation → Jobflows

Workspaces ​

  • Enables users to visualize and explore data already present in the File Explorer.
  • Supports:
    • Viewing validated datasets
    • Launching visualizations (charts, maps)
    • Previewing large files

Detailed documentation → Data Visualization and Workspaces

Members ​

  • Manage Project Space membership and roles.
  • Users with appropriate permissions can:
    • Add new members
    • Update member roles (viewer, editor, admin)
    • Remove members

Summary ​

Once signed in and inside a Project Space, you can:

  • Manage validated dataset templates
  • Upload, organize, and manage files
  • Launch and monitor jobs
  • Create, modify, and run routines
  • Build and execute complex jobflows
  • Explore and visualize data
  • Manage team members and permissions

This architecture provides a powerful environment for scientific data workflows, making Accelerator an ideal platform for:

  • Running complex model pipelines
  • Reproducing scientific experiments
  • Managing and versioning data
  • Sharing validated results