Inspiration
We knew that we wanted to do our analysis on a crime dataset because of the profound potential that it held. Any insights gained from a crime dataset could have a transformative impact. Inspiration struck when we found a dataset, brimming with possibilities, waiting to be explored. Our dataset was like a treasure trove of information and this inspired us to approach the data from a multitude of angles.
What it does
Our project is a dashboard that presents elaborate visualizations and insights. Not only does it have pre-defined visualizations, but it is also dynamic in nature. Users can modify the visualizations according to the month or even the year. We have made it a point to include numerous plots like barplots, piecharts and heatmaps
How we built it
In the initial phase, we began with performing intense EDA on our dataset. The formats for various columns had to be changed, the textual data needed to be sorted in categories and missing values had to be dealt with. Once we were happy with the state of our dataset, we began the process of deriving insights and converting them into visualizations. Each member of the team approached the data in their own unique way and soon we had multiple visualizations showcasing the potential of our dataset. Lastly, we started the process of building our dashboard using streamlit. Each visualization had to be categorized properly and then added onto to the dashboard.
Challenges we ran into
This is our first hackathon so we found it the most challenging to begin. Once we had started, the idea for the project struck us immediately but we kept running into errors after errors. The column formats were hard to change and we did feel dejected at times but we pushed through it and started to make the dashboard. The making of the dashboard was a challenge in itself. Everything had to be positioned correctly, the heat maps and bar plots had to be generated properly once the user provided input, and the insights had to be shown clearly. It wasn't easy but we're glad that we persevered and finished the project.
Accomplishments that we're proud of
We're very proud of the dynamic nature of our dashboard. We specially pat ourselves on the back for our creativity with the heat map and for actually making it work like we wanted it to. The heat map can even be generated for specific time-periods and this is what takes it to the next level.
What we learned
As beginners, the entire process was a learning journey for us. Working with such a big dataset and making the dashboard are all new milestones for the three of us. If we ran into errors, it took us a while to overcome them but I believe that's what made us truly learn new concepts. We enjoyed the journey and we'll be sure to continue enhancing our knowledge and performing even more intense data analysis.
What's next for Crime Severity Analysis: Assessing Impact on Victims
Next up, we're going to add some more categories for our dashboard with even more creative visualizations. And then we are going to challenge ourselves even more and make the dashboard so dynamic that users can choose which columns they want to see in the visualization. It will surely be difficult but we love a good challenge.
Log in or sign up for Devpost to join the conversation.