Georeferenced Program Evaluation (GPE)
Project led by urbanist data scientist Antonio Vasquez Brust, and described in detail in the IDB’s blog Behind the scenes: Big Data analysis for public policy (also available in Spanish).
The GPE (Georeferenced Program Evaluation) R package includes functions that will allow the user to study various aspects of consumer or beneficiary behavior, including:
Georeference postal addresses (converting street name and number to latitude and longitude points)
Obtain base maps of any city’s urban grid and visualize overlaid projections of the distance traveled (between participants’ homes and points of service accessed, for example)
Estimate a matrix of distances between sites accessed and consumer/beneficiary origin
Calculate metrics of frequency, distribution by group, and distance travelled by consumers/beneficiaries to points of sale/access
Create visualizations that explore the difference between frequency and type of consumption by consumer/beneficiary attribute (depending on the information available: socioeconomic status, length of program participation, type of program participation, etc)
Full code, description and installation instructions can be found here.