Peeing in the Sink
Skills
Data Science, Python, Seaborn, Numpy, Pandas, Data Visualization, Sci-Kit Learn
Objective
The age old question. Can I pee in the sink? Is it weird if I do? Some friends and I got in to a wager about whether or not people at the dorm where I live pee in the sinks in their rooms (by the way I don’t have a sink in my room) and it essentially grew into a survey and this analysis. I also wanted to practice using Seaborn.
Results
41% of the sample pees in the sink (Countplot)
2. American and Asian residents pee the most often in their sinks (Barchart)
3. Being further away from the bathroom makes you more likely to pee in the sink (Histogram)
4. Being religious means you have an even chance of peeing in the sink, whereas you are much less likely if you are not religious (Faceting)
5. Being younger makes you more likely to pee in the sink (Boxplot)
6. There isn’t enough data on people who don’t drink alcohol in the dorm, which matches the national average at 86% over the age of 18. Since N = 30, I could really only use logistic regression to see if I could predict whether or not people peed in their sink, whereas most other algorithms would overfit the sample.