For my final project of the Metis Data Science program, I investigated the climate change conversations taking place on Twitter in March 2017. The 1 million tweets that I looked at were a snapshot in time - many users talked about EPA head Scott Pruitt’s denial that CO2 causes global warming, or the critical condition of Australia’s Great Barrier Reef. Sub-communities were also apparent within the greater conversation, including a group of climate change deniers who stood out from the rest.

Here’s a simple dataset - just two variables that we’ll call \( x \) and \( y \).

The hardest part of an open-ended data science project is often defining concretely what your goal is.