TrafficData

Seattle Traffic Data Project

Purpose

The purpose of this project is to practice data science/engineering using Seattle traffic data.

Some of the longterm goals of the project:

• explore data science/software tools (AzureML, Jupyter Notebooks, Azure Function Apps, etc)
• be able to predict the evening commute hours in advance (eg: this evening's commute will be better or worse than average)
• look for trends over time on the different routes in the region

Plots

Jump straight to the plots here: TrafficDataViewer

Data

• WSDOT API: http://wsdot.com/traffic/api/

Components

TrafficArchiver

/src/dotnet/TrafficArchiver

This is an Azure Function App dedicated to retrieving the current travel times from the WSDOT Travel Times API every 20 minutes. It stores the data in an Azure Data Explorer (Kusto) table.

DataViewer

src/python/jupyter

A Jupyter Notebook for viewing plots of the data

Traffic

src/python/traffic

Some Python modules for getting data from the DB

DataArchiver

/src/dotnet/DataArchiver

This is an Azure Function App dedicated to retrieving the current travel times from the WSDOT Travel Times API every 20 minutes. It stores the data in a SQL table, also in Azure.

Setup

Requirements

Python 3.7

• pyodbc
• pandas
• matplotlib
• seaborn
• configparser

Create config file: TrafficData/_private/db.config

[CONNECTION]
server = <server>
database = <db>
uid = <uid>
dbpassword = <password>

Who

• Trevor Blanarik