Painstakingly creating a... Bandaid?
Hello, hello new friends! My name is Violet and I've recently fallen in love with data & data science! It started with a very hands on data coding course I took at ASU, and almost immediately turned into a real world project, to help Reverie Books, a lesbian owned bookstore, overcome the difficulties with using Square to sell books.
You've probably been to a food truck that used Square, and it works great for that application. However, it isn't quite the same experience using it for a bookstore. In order to put together all the data they used to decide what books to order took 3 spreadsheets exported from Square, and about 5 or 6 hours of wrangling it. Surely, in current year, it shouldn't have to take that much time to wrangle data for a weekly book order.
I started this project early into the semester in my data coding course. The course started with wrangling data with SQL and then moved on to using Python after we understood the basics of data wrangling and joining/filtering/etc of multiple tables. When I started working with Reverie, we were still in the SQL part of the course and I wanted to use my new skills to tackle this problem. So I made a database, shoved the spreadsheets from Square into it, deleted and remade the tables to fix data type issues more times than I'd like to admit. Once I got the tables in a workable state, I was able to write a query to get just the information we wanted to make ordering decisions. This felt like a successful proof of concept, but spending hours hammering spreadsheets into a database in order to save time on data wrangling didn't really feel like a solution. I wanted to make something that was usable and wouldn't have a steep learning curve.
Around that time, my course moved on to the second part, which was using Python and Jupyter Notebooks to manipulate data. I went out looking for a free Jupyter like solution since I didn't want to do this in an environment where my access was tied to a course that would end. So I ended up in Google Colab, which worked just as well albeit with a side of "I hope they don't randomly delete this product". So I took what I learned from using SQL to wrangle this data and started digging in with Python. I've worked with Python on and off for years so it was nice to get back to what I'm used to. Pretty quick I was able to get the data in and start manipulating it. There was also talk of fixing book categories and calculating turn over for each category, but getting all that sorted found its way to the back burner to make way for streamlining the process to get the data to order by.
Getting the notebook to provide the data we wanted in an intuitive way gave me a lot of experience wrangling data and a glimpse into the monster that is Excel. After multiple iterations of wrangling to get the correct data and displaying it in the best way to make ordering decisions, I'd finally done it! I've created a bandaid that bridges from now until Square rolls out their fixes to make their platform more hospitable to non-foodtruck uses. Now in exchange for the 3 spreadsheets from Square, all that you need to do is put the in the folder with the notebook and click run all cells. A couple minutes later and you get back a spreadsheet with the data to make informed ordering decisions, and not any extra nonsense like if each transaction was dine in or to go.
This project all started with me asking the owner if she had anything that a data science student could help with. I truly had no idea what that might end up being when I asked, but I'm so glad I did. Working with Reverie has been a treat, and taught me a lot about working with small businesses and given me the courage to put myself out there. The course I took was very hands on and did a fantastic job to prepare me for a real world project before the semester was even over. I was quite familiar with Python and had used SQL once or twice going in, but it really helped to put a functional edge on that experience and direction for my career.
Since that semester, I've accepted a job being a Team Lead for the very course that sent me down the data rabbit hole. My first live lab is coming up next week, and I am so excited to share the amazing experience with new students!