Earlier this year, the Manhattan Borough President's Office (MBPO) conducted a survey of vacant storefronts along Broadway in Manhattan. Using the borough's longest street (and the only one to go from the very top to the very bottom of the island), as a representation of the trends across the island, the MBPO put together a comprehensive report.

While the report only lists the addresses of the vacant buildings, I thought it would be interesting to map the locations out and see if any neighborhood-level trends emerged. To do this, I took advantage of the bokeh python library, which is one of my favorites for data visualization. Using bokeh's Google Maps plotting tools and the geopy geocoding library from python, I was able to read each address and map it to where it falls along Broadway.

The full code I used to make the map as well as the final output is shown below, courtesy of Kyso, a new software platform for sharing data science projects - specifically projects developed in Jupyter notebooks.