San Francisco Crimes time analysis and GeoData

• Tech Stack: plotly, folium, matplotlib, seaborn, Pandas
• Embedded Notebook: Link
• Github URL: Project Link
• Contributors: Peter Anthony Wright and Carlos Marcos Torrejon

The San Francisco crime dataset from DATASF provides a special opportunity to play with data visually and try to understand the underlying patterns present in it. In this project, we only focus on a subset of the crimes category. Since the dataset covers a wide variety of crimes, visualizing them all at once might wash out any patterns contained in smaller subgroups Our Analysis is guided by a Consistent Visual Platform which we use to direct the path of interest of the User, providing multiple analysis as well as summaries and annotations. We also decided to realise many of our graphs interactive using several libraries form in order to deliver a more stimulating experience for the whomever is interested in visualising our work, even if it is not their area of expertise.

Possible Optimisation

• Use Spark for data processing
• Write python code in OOP style instead of in Notebooks. Improve in organization, reutilization and flexibility, and scalability.