Session 1: Tools - Setting up Python and R in Classroom
Presentation Type
Event
Abstract
Python and R are perhaps the two most popular free analytics platforms for Data Science. Unfortunately, for novice users such as students, installation of proper components and tool configuration is usually a problem. Even more difficult it is to ensure that everyone in the classroom works in the same environment, uses the same package versions, etc. Here, we present a classroom setup, where the students access their Python and R notebooks through HTTP via JupyterHub server. In this setup, the students have full permissions to their home directories as well as read/execute access to shared notebooks and data directories. The instructors can use the latter two directories to share in-class work, lectures, and data with all the students. The JupyterHub server runs on a virtual Linux machine to facilitate resource management and backups. This approach has been tested on smaller classes of up to 15 simultaneous users (both students and faculty). For larger classes, JupyterHub can be deployed on multiple nodes using Docker Swarm. Overall, this setup is an excellent platform where the users can focus on their work: learning or research without wasting time on package configuration, backups, and resource management.
Start Date
2-12-2018 11:00 AM
End Date
2-12-2018 12:00 PM
Session 1: Tools - Setting up Python and R in Classroom
University Student Union: Clark Room 262 B
Python and R are perhaps the two most popular free analytics platforms for Data Science. Unfortunately, for novice users such as students, installation of proper components and tool configuration is usually a problem. Even more difficult it is to ensure that everyone in the classroom works in the same environment, uses the same package versions, etc. Here, we present a classroom setup, where the students access their Python and R notebooks through HTTP via JupyterHub server. In this setup, the students have full permissions to their home directories as well as read/execute access to shared notebooks and data directories. The instructors can use the latter two directories to share in-class work, lectures, and data with all the students. The JupyterHub server runs on a virtual Linux machine to facilitate resource management and backups. This approach has been tested on smaller classes of up to 15 simultaneous users (both students and faculty). For larger classes, JupyterHub can be deployed on multiple nodes using Docker Swarm. Overall, this setup is an excellent platform where the users can focus on their work: learning or research without wasting time on package configuration, backups, and resource management.