#un general

rishsriv:

Based on data from the UN voting records, compiled by Erik Voeten and Anton Strezhnev here. Data processed using pandas and plotted using Google Geocharts.

To calculate similarity of voting, the following steps were taken:

A "Yes" vote was given a numerical value 1

A "No" vote was given a numerical value 0

The options of "Abstain", "Absent", and "Not a member" were ignored in the analysis

If both the US and a given country had voted on >5 resolutions between 1992 and 2012, a pearson correlation coefficient was determined for the votes of the two countries. This was then used as a metric for comparing similarity.

Correlation coefficient was used as a metric instead of cosine similarity or jaccard similarity for ease of understanding. I'm not entirely sure if that's a wise choice though. Criticism appreciated!