Wardrobe Database

Project Introduction

Pareto's principle has long been a fascinating subject for me. More commonly known as the 80-20 rule, this paradigm is usually applied to business or economic examples, however, it's been proven to relate to all facets of life.

In looking at my personal life for applications and opportunities to envoke Pareto's principle, I found an example in a seemingly unlikely place: my wardrobe.

If you're anything like me, you wear the same "x" number of shirts and "y" number of pants, default to the same rotation of shoes and accessories. There must be an anthropological or psychological explanation for such a phenomena.

Furthermore, I've noticed most of us continually add to our clothing collection and rarely takeaway (talk about a nightmare inventory and accounting system were this a business).

This was a real pain point for me, and likely for others, which means it's an opportunity for technology and data.

Situation

After noticing that I cycle through the same articles of clothing, I wanted to better align my habits and preferences with the phsyical composition of my wardrobe.

If I wear a lot of "green" colored items, shouldn't I have a similar proportion of items in my collection? Furthermore, such a tool has the potential to help determine if a new item of clothing is likely to be worn enough to merit the purchase.

Ultimately, this project is about efficiency in an area I'd wager most are worse off than they think.

Task

In order to capture the insights related to dressing habits and wardrobe preferences, we need a powerful data source.

This source would need to catalog an entire wardrobe with features such as color, patterns, materials, etc. Furthermore, it'd need to also document the outfits worn and for what occasions. Beginning also with the end in mind of such a data source extending beyond just myself, we'll want to associate outfits and their associated articles of clothing to specific owners.

I want this data source to be accessible to other programs for analysis and visualization, but also maybe one day to be powered by a front end application. For now, however, we'll concentrate on the data source.

Action

As detailed extensively in the project components below, I took steps to design the database structure to afford an efficient and yet detailed cataloging of wardrobe and outfit data. I really wanted this project to feature as much practice with databases as setting up the perfect source.

To that end, I created a PDF schema documenting the intricacies of the database. My experience in college didn't go into creation of schemas so I researched on my own and am continually updating where needed.

I also created scripts, designed functions and procedures to create the database. I have plans to deploy it to the cloud, likely using Digital Ocean as the host.

I plan to load the wardrobe data and document outfits over the duration of a year. At that point, I plan to access the data with R or Python for exploratory data analysis, machine learning, and advanced visualization.

More opportunities will surely unfold over time, such as making a portion of the database available to the public via a dataset and creating a front-end program for easier data entry.

Result

I'm extremely happy with the early results! From the structure and integrity of the source to the ease of accessing the data via R and Python. There's much more to come but take a look at the project components to get a feel for the database.