Posts Tagged “python”


Gun Violence Incident Grouping via PCA

Note (Oct. 14, 2016): Part of this work was used in a collaboration between the AP here and USA TODAY NETWORK here.

The Gun Violence Archive (GVA) has for each event a list of characteristics. There are 92 distinct tags that are used. Many of these tags are either redundant or unnecessarily specific for the kinds of questions I am interested in investigating:



Gun Violence Statistics

Tags: python d3js sql

The Gun Violence Archive (GVA) is a database of gun violence and gun crime in general aggregated from a diverse array of sources. The specific details of their methodology for tagging incidents is listed under their Methodology.



Dota 2 Drafting Part 2: Hero Pairings


The data collected from Valve’s API contains a plethora of information about relationships between heroes, teams, and winrates. This data lends itself to the construction of graphs depicting the relationships. This has a twofold purpose:

  1. To assess the integrity of the collected data. If the graphs that come out don’t line up with the common strategic thinking about the heroes then additional stratification is necessary (most likely by game skill bracket).
  2. To explore the data set to see what hero combinations are popular along with showing which combinations work well and which don’t.



Dota 2 Drafting Part 1: Data Collection

Dota 2 captain’s mode entails a drafting phase where a sequence of picks and bans are issued by the team captains. As of this writing, there have been 1,933,725,512 total matches of Dota 2 played, in all modes. It may be possible now with this body of data to build a machine learning algorithm for prediction and for drafting suggestions.