Platform Introduction Video
Codebench is an integrated platform for learning data science through Jupyter notebooks. It assists in desigining, developing,releasing, and overseeing project submissions. During the assignment completion phase, users can leverage generative AI for responsible AI assistance. It is a complete pipeline making it easier to manage data science and AI courses.
Learn more about CodeBenchSee below for comparisons with standard localized coding environments and google colab - typically used in Data Science and AI courses.
Coding Environments
The codebench is rich with many built-in coding environments from - command line, python, R, Java, Matlab, Spark allowing codebench to be used in many Data Science/AI course as well an instructional tool by instructor/TA's for class demos as well as students using it for completing assignments, receiving help, submitting and feedback from graders. The table below provides all coding environment where one can use Jupyter notebooks or command line for developing, testing and submitting code. Codebench requires no configurations by the student and core libraries are included in the distribution. Each student can install their own libraries as needed.
Learn more about CodeBench See below for comparisons with standard localized coding environments and google colab - typically used in Data Science and AI courses.
CodeBench vs Google colab vs Canvas
Unlike standard tools used in data science education such as local installations and Google colab, codebench is designed as a course workflow based architecture. Here we compare codebench to other methods.| Task | Using Codebench | Using Canvas | Google Colab |
|---|---|---|---|
| Installing python or R-studio | No installations necessary | Must instruct students to install python or R locally | Python is available |
| Coding on command line | Available | Not Available | Not Available |
| Access to Jupyterlab | Access on the cloud/server. All libraries installed by admin. Same coding environment for all. | Students must install JupyterLab locally. Each configuration is time-consuming. | Accessible on the cloud with a personal Google account. Must configure necessary libraries. |
| Handling Large Data Files | Centrally managed. Easy to update and access shared data. | Each user must locally install. Hard to update large data once released. | Medium data via Google Drive, large via Google Cloud (costly). |
| Preparing assignments | Built in cloud/server for testing and development | Developed locally, uploaded to Canvas | Developed on Colab, then downloaded/shared via Canvas |
| Releasing Assignments | One-click release to students | Uploaded to Canvas, students download locally | Notebook shared, students make a copy to complete |
| Completing assignments | Students complete in the cloud/server Jupyter environment | Completed locally in Jupyter Notebooks | Completed in Colab in student accounts |
| Submitting assignments | One-click submission, multiple submissions allowed | Students upload to Canvas | Download Colab and upload to Canvas |
| Late submissions | Automatically handled | Manually handled | Manually handled |
| In Class Coding Demos | Built-in Capability | Not Available | Available |
Universities and Colleges
Rutgers University
Raritan Valley Community College
Cost of Use
Codebench was developed at Rutgers University - New Brunswick and is governed by the IP laws of the University. It is provided for free or low cost to Community Colleges and Institutions for use in their data science courses. Please contact codebenchai@gmail.com for more information.Connecting to Canvas and CAS
It is possible to connect CodeBench to Canvas or use CAS for common authentication. Students can use CodeBench with their institutional email.Try It Yourself
Access the public instance of CodeBench to explore its features and experience AI-powered coding education firsthand. Please fill out this form to request access