For my final project in the Metis data science bootcamp, I chose to investigate HIV and AIDS data provided by the United Nations for their Millennium Development Goals data visualization challenge.
You can see the finished project on my website at emilyschuch.com/UN-challenge
- 01_UNreshape: reads in the data from a Postgres database, unstacks the series I've chosen to work with, and merges in some additional variables downloaded from the World Bank.
- 02_UNexplore: uses plotting functions to look for trends among the available variables.
- 03_UNexplore: reshapes the data further, investigates percent change of some variables from previous year.
- 04_UNmodel: explores various predictive models, both linear and classification.
The three main visualizations are:
- A line plot of the HIV incidence rate by region (un-line.js)
- A world map of the HIV incidence rate by country with a slider to view by year from 1990 to 2013 (un.js, d3.slider.js)
- A heat map of the 20 countries with the highest average HIV incidence rate from 1990 to 2013, by country and by year (un.js)
- Amazon Web Services
- SQL
- Python (pandas, statsmodels, sklearn)
- Linear regression and classification modeling
- HTML / CSS
- D3.js