cs6141-final-project 0.1.2

To infinity, and beyond!

CS6140-final-project

Authors: Daniel Blum, Matthew Greene, Bram Ziltzer

Summary: A classification project using Logistic Regression, SVMs, and DNNs to predict if a flight will be delayed or not.

The Docker Python Environment

The project is running in a collection of docker images via docker compose. The jupyter-lab image runs a Debian instance with Mamba as the environment manager.

Running the Poetry Environment

Prerequisites:

• Docker
• Something to run Jupyter Notebook/Lab UI like VSCode or DataSpell

1. cd into the parent directory of the project
2. run docker compose up -d

Docker is going to do its thing for a bit and download a good amount of stuff. Once done, you should have control over the terminal (as the -d is the detached option for docker compose).

If you run docker ps, you should have these running services:

• jupyter-lab

And a new network with docker network ls:

• cs6140-final-project_cs6140-network

Updating Poetry

If you need to add libraries TODO (dan) - FILL IN

Updating the PostgreSQL Image

Running Jupyter

VSCode

Prerequisites

• Jupyter Extension for VSCode
• Python Extension for VSCode

When you open a *.ipynb file in VSCode, you'll want to the remote kernel running in docker.

1. In the top right corner of the ipynb file, select Select Kernel (if you don't see this, it might have another kernel already selected)l.
2. A drop-down pops up in the center of VSCode, select Select Another Kernel.
3. Select Existing Jupyter Server...
4. Enter http://127.0.0.1:8889
5. Select Python 3 (ipykernel), the kernel at /mamba/bin/python

You should be all set!

Browser

Prerequisites

• A Web Browser

Open up http://localhost:8889

DataSpell

Running Poetry

don't worry about this