Welcome To Denver Dryver

Liking the flexibility that rideshare platforms offer. I decided to take up rideshare driving while I was taking classes in the spring semester of 2017. With ok profits resulting from it. i kept driving in the summer where I discovered that the weekends are a great source of revenue. Realizing this I started driving full time in the summers. I did this for two years in 2018 and 2019. I mostly drove for Uber with Lyft as a sidekick when things got slow. In total, I drove 3200+ trips with Uber and 300+ with Lyft. I unknowingly created a small data set that was too good to forget, so I didn't. I wrote some web scrapers to collect my driving information and processed the data which can be seen below.

Uber makes it very hard to collect trip information. Each trip has its own HTML and statements data ordering is inconsistent across months. Thus the first challenge was data collection. Test bots were created with Beautiful Soup and Selenium. In the end, I decided to go mainly with Selenium as the HTML code on Ubers site was too dynamic and sloppy to use. With Selenium, I was able to collect the vast majority of my trips on the platform. The data was processed after collection and combined with statements to answer as many questions as possible. Pandas, Python, Excel, Juypeter Notebooks, & Visual Studios Code were used for this step, which was the most time-consuming aspect of the project. The final step in the project is constructing this website which, was wireframed on paper. The website itself was built from scratch using HTM5, CSS3, jquery, BootStrap 4, PHP, Charts.js, Typed.js, & Counter.js. It took one month of work to get to this point and not all the questions I answered with the data set are posted on this website.

One of the main drawbacks is that it only my data set. 35-40 hours spread over Friday, Saturday, & Sundays in the summer months mostly. Thus I am missing the weekdays and 8 non-summer months. Collecting other driver data would be a great solution to this problem but I would need account access for my web scrapper, and would potentially compromise their income source. I also don't know that many Uber drivers, so we'll have to settle for my data set. In a nutshell, I think that this is a great data set, but it's far from big enough to take extremely seriously. The data is presented as is to the best of my data science/ data mining abilities. All this data is pre COVID-19, so things may be different out there.