Pirate Attack Probability

Inspiration

Our inspiration was the theme which was about pirates so we choose to make our gui and product pirate centric.

What it does

This uses a large dataset from Kaggle, which gives a large variety of information regarding pirate attacks from 1980 to 2020. When run, a GUI pops up and asked for a latitude and longitude, and once they have been fed in, a probability is outputted.

How we built it

We built it using Python, using "tkinter" for the GUI and then calculating the chance of being attacked by pirates by doing a linear regression-adjacent formula coupled with additional manipulation to account for faults in the data set.

Challenges we ran into

The data set was very precise so finding correct values to represent the actual probability was difficult, making the creation of a formula from scratch challenging. Additionally, some points were extremely spread out form each other, and the system we used of finding the slope between two separate lats/longs and their attack numbers was also a little skewed. Originally, it showed a higher chance of being robbed in the middle of the Atlantic ocean than right next to the southern coast of India, where pirates exist in concentration. This was solved by accounting for the difference between the two points the slope was made from and adding it into our formula calculations.

Accomplishments that we're proud of

We are proud of the GUI and the way we were able to integrate that with our backend. Additionally, the probability is accurate for most locations, and we're glad that the formula works.

What we learned

We learned a bit about using extensions on VS Code to collaborate; we also learned more about how Python works and about some libraries, like Pandas. Additionally, we learned how to collaborate with LiveShare on VSCode as well.

What's next for Pirate Probability

Our next steps are to increase the accuracy of the probabilities when a long./lat. on land is chosen: As the current program calculates the probability of a pirate attack on specific latitudes and longitudes and multiplies them together, if a point on land lies on a long./lat. with a lot of pirate attacks elsewhere, it may incorrectly identify that location as one of high risk of an attack.

Built With

• python
• tkinter