Locatelli Photography

Code

 

Click that thing up there, that’s my GitHub. There’s not a lot on it right now, but there will be in the near future.

 

A growing amount of my work involves writing scripts, primarily in Python and in R. One of my biggest pet peeves is data accessibility and the availability of easy to follow tutorials that can be useful to both beginner and advanced coders. As such, I’ve made it my goal to make all scripts I write during my academic career - both for data analysis and data visualization - publicly accessible on GitHub or whatever platform may replace it in the future.

My frustration with data availability began during my final project for the Data Science and Programming for Biologists course at Columbia University, taught by Deren Eaton. For my project, I sought out to create an interactive map of over 10,000 coral reefs worldwide called VirtualDive (see the unfinished but pretty map below and the repository for how I put it together. If it doesn’t load, try refreshing the page. Sometimes it gets stuck when switching between pages on my website). My goal was to create an exploratory tool for people to use for fun or for choosing research sites based on reef characteristics such as community composition. To create this map, I used Bokeh, a library which takes input in Python and outputs data visualizations using a Javascript API. This ended up being a lot to chew, since I had only just begun learning Python. During the creation of VirtualDive, I spent hours trying to find answers to questions that could easily be resolved in minutes if detailed annotated code from other people was publicly available. The community of Bokeh users seemed to consist largely of either people who were experts or people who knew nothing, but no one of intermediate knowledge to bridge this disconnect. As a result, the purpose of my project pivoted from creating a polished tool and instead towards creating a heavily annotated tutorial that utilized Jupyter Notebooks. The goal of creating these annotated scripts was to allow for anyone of any knowledge level to begin using Bokeh and to increase the accessibility of this powerful visualization tool and the findings that would be presented using it.

 Data for VirtualDive from ReefBase Reef Locations dataset. Warmth of colors on land indicate areas of high human population density.