Backend for Yelp Data visualization application.

Steps below are from the official Flask documentation. Setting up a virtual environment and using the requirements.txt is highly recommended. This prevents package confusion that frequently occurs with global packages and will make it much easier to debug problems with pip on your computer.

virtualenv venv

On OS X, do the following

. venv/bin/activate

If you are a windows user,

venv\Scripts\activate

To deactivate the virtual environment:

deactivate

After activating the virtual environment, run the following commands:

sudo easy_install pip, if you do not already have pip (you usually should).

pip install -r requirements.txt NOTE: when adding new packages, run pip freeze > requirements.txt after to add to requirements.

To run the server, run python app.py and go to localhost:5000 in your browser.

For Mac users: brew install mysql brew services start mysql

For Windows users, taken from the MySQL site:

1. Download the MySQL Installer and execute it.
2. Choose "Developer Default" as your Setup Type.
3. Follow the instructions and complete the installation.

The next instructions are for setting up the database and user account. If you are on Windows, the steps could be done through MySQL workbench (without running any commands). If you are on Mac, simply run the commands.

Set password for root account, mysqladmin -u root password 'poloisbae'

Connect to MySQL using username root and password poloisbae, mysql -uroot -ppoloisbae

Once connected to MySQL, create a database with the name "db" create database db;

To import the data, download the Yelp dataset from here: https://www.yelp.com/dataset/download

Once downloaded use this command to import the data into the database (may take 5-10 min to import):

mysql -uroot -ppoloisbae db < yelp_db.sql

Download the cosine.sql and index.sql here which contains the result of training on 10605 restaurants:

business_cosine: https://drive.google.com/file/d/1nyuksX9Ht942XfFJxM1gdNX24hUh2qCl/view?usp=sharing

business_index: https://drive.google.com/file/d/1paildYInWCSlbIIt4Uvq0c0uX1R09Iox/view?usp=sharing

Import the table business_cosine with mappings from business id to cosine vectors using:

mysql -uroot -ppoloisbae yelp_db < cosine.sql

Import the table business_index using:

mysql -uroot -ppoloisbae yelp_db < index.sql

This table is used to determine which indices in the cosine vectors map to which businesses.

Run python app.py

Use postman to try a sample query:

Use this post data as json for example:

{"sources": [{"id": "-01XupAWZEXbdNbxNg5mEg"}], "exclude":[]}