Coding Projects

Summary

This project is my largest ongoing project. It started as a solution to the lack of real-world data available about all the things that go on in an economy. The overarching goal of this project is to have the capability to generate realistic data of everything that happens in an economy and use that data to interpolate and analyze results to product price changes and new product introductions. The program initializes by creating a desired amount of consumer Artificial Intelligence (AI) and a desired amount of Business acting AI. Consumer AI's are trained on real world data about propensity to spend by industry, income, average savings rate by education level. Each AI is giving a certain amount of income, financial assets, debt, wants and needs. Each Consumer AI is then released into a continually evolving economy that is composed of many businesses selling multiple products. Products have 10 different attribute values that correlate with 10 statistically-generated characteristic values that can sway a consumer to purchase a product. The product attributes determine how expensive it is to product the product and will determine how much profit each company makes.Businesses vary certain expenses to maximize profit or get the product in front of as many consumers as possible. At the end of each fiscal year of the simulation, companies will add/modify/remove products or will go bankrupt.

Summary

What originally started out as a test of my web-scraping skills, evolved into a much larger project. Every 2 hours, my custom web-scraper goes and collects 124 datapoints for every location in the United States. This works by making thousands of requests to publicly available geographic image layers and extracts and interpolates the values. This project was composed of many web-scraping components as well as many image-processing and manipulation techniques that I learned specifically make this work. The goal is to collect a years' worth of data and package and sell it, acting as a Data Broker to companies and anybody who wants the historical data.

Summary

While enrolled in my Web-Scraping class during my Master's program at the University of Florida, I decided to try my hand at fully automating my multiple online businesses so that way there would be no human input required to run. As of writing, I have 3 online Etsy businesses that sell print-on-demand images that require no physical touch to product and ship. When first starting these businesses, I would hand print and ship every order. Eventually I would find a print-fulfilment company that would automate all of this for a small fee. This still required me to input the file, address and payment information. My goal with this project was to fully automate this process. I wrote a program that completed each of the below steps: 1. Check for any unfulfilled orders 2. Collect information for orders 3. Select and resize the purchased image 4. Input image, address and payment information into print-fulfilment website This project is completed and executes 4 times a day to automatically forward orders to the print fulfilment company. It has done so flawlessly for over 200 orders across my multiple online shops.

Summary

For my third online automated business, I wanted to bring many of my skills together of, web-scraping, image-manipulation, data-mining & cleansing as well as incorporating my previous Automated Online Store code to run the business side. This was my second largest project and could be broken up into 3 large sections. The goal of this project was to web-scrape athletes names and then arrange those names in the shape of the logo of the team that they play for. Multiple variations of the team's logos would be created from the list of player names 1. Web-Scraping Athlete Names -I mulitple websites that had thousands of athlete names for each time. In total I scraped over 114,000 individual names from Professional Football, College Football, Professional Basketball, Baseball and Hockey. - External data was used to determine the size of each athlete. The more recent an athlete played for a team as well as their prominence in the news, determined their size in the final image. 2. Wordcloud Creation - I used, modified and tunes a python package to create the wordcloud images of each team logo. This created over 400 individual images for over 163 teams. 3. Mockup Creation -Once I had code that created the final images. I needed a way to upload them to my Etsy store. In my prior two online stores, I painstakingly uploaded each image and wrote each description. With this project, I wanted to do all of this automatically so I could scale to more and more teams if I wanted to in the future

Summary

For Spring Break 2022, my family and I went to Washington DC together. While there we toured many of the museums and Smithsonians on the National Mall. While there we Hirschhorn art exhibit and I was able to witness Reynier Leyva Novo's, The Weight of History: Five Nights installation. It was simply composed of 5 black rectangles of varying sizes hanging on the wall. At first, I found it extremely comical but upon further investigation, I realized that each of these 5 squares represented the amount of ink used to print fascist publications throughout history. I found this concept to be really interesting so I set out to build a lightweight program that would automatically calculate the size of the black square to represent any input text. This program took an input text, converted each letter to an image, counted the number of non-white pixels and finally summed up the results into an appropirately sized rectangle.