Sea To See

Inspiration

The theme of the event pushed us to dive into the idea of exploring the unknown of our Earth's future. Therefore, we decided to look into the realm of sustainability, and we thought trying to parse data from the fishing industry would be a good way to map out how humanity could better comprehend potential outcomes from what is already known.

What it does

The back end was mostly numerical computations. The language of choice was Python, with SQL also used to query the data set and organize the data. Most of the Python used numpy for numerical computations, matplotlib to create the graphics, and special parts of statstools package to generate time series models. In particular, the type of time series model generated was an ARMA model (autoregressive moving average). This forecasts based on previous data and noise terms that are averaged out. For the front end, our project displays a static web page created with Bootstrap, HTML, and CSS that shows our scatterplots and ARMA models as png files along with supplementary information.

How we built it

We used a collaboration of Google Colab and VSCode for the backend side of the project. The frontend used Adobe Dreamweaver to create the webpage & configured an AWS web server.

Challenges we ran into

We had issues initially with getting version control to work with Google Colab and GitHub, so we then used a combination of GC and VS Code. We also had issue with implementing the flask library as a means to transfer matplotlib graphs into png files.