Surgeo

The documentation for Surgeo may be found here: https://surgeo.readthedocs.io/en/master/

Contributors

• Adam Weeden
• Algorex Health
• Theo Naunheim

Overview

Surgeo is a module that contains a variety of open source demographic tools that allow you to construct race probabilities from more commonly available information such as location, first name, and last name information. This imputed race data is often used in the public health and fair lending contexts when race information is not otherwise available.

Specifically Surgeo contains the following models:

• Bayesian Improved First Name Surname Geocode (BIFSG): an adaptation of an algorithm created by Ioan Voicu that uses forename, surname, and location information to obtain probable races
• Bayesian Improved Surname Geocode (BISG): an adaptation of an algorithm created by Mark Elliot and popularized by the Consumer Financial Protection Bureau (CFPB) that uses surname and location to obtain probable races
• Forename: a helper model to pull race data based on first name
• Surname: a helper model to pull race data based on last name
• Geocode: a helper model to pull race data based on location

Please see the ReadTheDocs link above for additional information on the data sources used and the implementations themselves.

Installation

To install surgeo as an executable, please see the installer below.

To install as a Python module, you can use pip:

$ pip install surgeo

Usage

Surgeo can be used as a stand-alone executable or a Python module. Details follow.

As a Program

To use the GUI, simply type in "surgeo_gui" or use the Start Menu after installing the executable. For Mac or Linux users, ensure that you have tkinter setup on your Python distribution.

$ surgeo_gui
# Or alternatively if you have installed the module
$ python -m surgeo

To use the CLI, type in "surgeo" followed by your arguments.

$ surgeo_cli --help
# Or alternatively if you have installed the module
$ python -m surgeo -h

usage: cli.py [-h] [--zcta_column ZCTA_COLUMN]
[-ct]
[--first_name_column FIRST_NAME_COLUMN]
[--surname_column SURNAME_COLUMN]
[--state_column STATE_COLUMN]
[--county_column COUNTY_COLUMN]
[--tract_column TRACT_COLUMN]
input output type

Get Surgeo arguments.

input                 Input CSV or XLSX of data.
output                Output CSV or XLSX of data.
type                  The model type being run ("first", "sur", "geo", "bifsg", or "surgeo")

optional arguments:
-h, --help            show this help message and exit
-ct                  Process for CENSUS Tract as opposed to ZCTA/ZIP
--zcta_column ZCTA_COLUMN
          The input column to analyze as ZCTA/ZIP
--first_name_column FIRST_NAME_COLUMN
          The input column to analyze as first name
--surname_column SURNAME_COLUMN
          The input column to analyze as surname
--state_column STATE_COLUMN input column containing two digit FIPS state code
--county_column input column containing three digit FIPS County Code
--tract_column input column containing six digit tract code

As a Module

Surgeo is best used as a module.

import pandas as pd
import surgeo

# Instatiate your model
fsg = surgeo.BIFSGModel()

# Create pd.Series objects to analze (or load them)
first_names = pd.Series(['HECTOR', 'PHILLIP', 'JANICE'])
surnames = pd.Series(['DIAZ', 'JOHNSON', 'WASHINGTON'])
zctas = pd.Series(['65201', '63144', '63110'])

# Get results using the get_probabilities() function
fsg_results = fsg.get_probabilities(first_names, surnames, zctas)

# Show Surgeo BIFSG results
fsg_results

Prefab Files

A link to the Windows GUI/CLI is below.

Windows installer.