We were inspired by the data set provided by NCR to build a data visualization of a stores best selling departments in different cities across the nation. We also build a simulation of shoppers to analyze the efficiency and shopper-friendly design of the stores locations per city. We wanted to create a analytical view of the entire lifecycle of selling a product, from warehouse distribution to product placement within each specific store.
What it does
Our main page shows the most popular locations of a store franchise based on the number of transactions. When hovering over a city location out donut chart automatically updates to represent the percentages of departmental transactions for that specific store location. When the city location is clicked the user is brought to a simulations of shoppers experiences buying their items. This tool can be used to analyze the shopper-friendliness of the stores layout based on shopper traffic while picking up their items. It also simulates the restocking of the departments to provide an even deeper analysis of popularity of certain departments and their items.
How We built it
The back-end was build with Python and Flask. We loaded the csv data with an added column of locations from NCR and stored it in an SQLite database to support out custom API. The generalized data for the nation and city is created using D3.js for data visualization. The store analysis was made using Unity.
Challenges we ran into
Some challenges we ran into were the layering of SVGs to create meaningful representations of data pulled from our API. Integrating all of the different frameworks we used.
Accomplishments that we are proud of
Creating a custom API. Creating meaningful data visualizations with D3.js
What we learned
We learned how to use Flask and D3 to make a custom API and data visualizations.
What's next for TBD
Integrate warehouse integration and analysis for end-to-end optimization analysis.