najah-lshanableh/lead-public

DSaPP Lead Hazard Modeling

CDPH Lead Hazard Model

This code models lead hazard in the city of Chicago. This project is under development by Eric Potash and Joe Walsh at the University of Chicago's Center for Data Science and Public Policy in partnership with the Chicago Department of Public Health.

Closely based on previous work of Joe Brew, Alex Loewi, Subho Majumdar, and Andrew Reece as part of the 2014 Data Science for Social Good Summer Fellowship.

The Problem

About 90% of Chicago's housing stock was built before 1978, the year that the Federal Government prohibited almost all lead from paint. Lead poisoning affects neurological development, leading to lower IQs, increased need for special education, lower wages, and increased crime. Lead has the greatest effect during the first two years of the child's life, but many kids don't get tested until later.

The Solution

The pipeline consists of three phases which are summarized below. The code for each phase is located in the corresponding subdirectory. The output of each phase is contained in a database schema of the same name.

##input Here we preprocess and import our data into the database. CDPH provided us with two databases about lead:

• Blood Lead Level Tests
• Home Inspections

We supplemented that data with the following public datasets:

• Chicago addresses
• Cook County Assessor Data
• Building
• American Community Survey

##aux Here we process the data to prepare for model building. TODO details.

##output Here we assmeble the data for modeling. TODO details.

##model Here we run models using GNU Parallel and pickle the results.

Running the model

Drake controls the workflow.

Install software

• Drake
• mdbtools: a command-line tool that reads Microsoft Access files, which is useful for reading the property-assessment data