saurabh21289 / Food-and-Stuff

A full-stack restaurant suggestion system using machine learning concepts

Food and Stuff (EC601)

Demo v0.4 is live at http://saurabh.tech:4000

Technologies used are Python 2.7.12 with SQLite, Mrjob, Flask_googlemaps, Pandas, NumPy, Flask with NGINX, Html5, JavaScript, Bootstrap, Jinja2, Google maps API, Charts.js, SQLAlchemy

How to run:

The project files are inside the "ML Engine" folder.

Requirements:

1. python 2.7.12 with sqlite

2. Extract dataset file (linked_cate_name.json) inside the "ML Engine Folder"

Link to dataset: https://github.com/saurabh21289/EC601-Food-and-Stuff/raw/master/linked_cate_name.zip -or- https://www.dropbox.com/s/v8ko9iy8ot7in7f/dataset.zip?dl=0

Note: The zip file above contains "linked_cate_name.json" which is our pre-processed dataset. Ensure that you place this file inside the "ML Engine" folder so that server.py can locate it.

Compatibility: Windows 8.1/10 and Ubuntu 14.04. Also tested on macOS.

Step 1: Setup python environment (may need sudo on Linux)

pip install Flask-SqlAlchemy wtforms flask flask_googlemaps

Step 2: Setup database (may need sudo for Linux)

python tabledef.py

Populate database with dummy data (may need sudo for Linux)

python dummy.py

Step 3: Running the application without NGINX

For both Windows 10 and Linux:

python server.py

Tip: If want to use NGINX, install uWSGI using sudo pip install uwsgi. Then run:

./start.sh

The server runs on http://localhost:5000 by default. If you want to host on a different port, change inside server.py for Windows/Linux and in start.sh for Linux with NGINX.

Feel free to reach out at [e-mail] if you have any trouble running this application.