PopGen3

Introduction

PopGen is a state-of-the-art open-source synthetic population generator for advanced travel demand modeling. Developed under the guidance of Professor Ram M. Pendyala at Arizona State University, PopGen utilizes a heuristic algorithm (IPU) to control and match both household-level and person-level attribute distributions.

Key Features

• Advanced control of household and person variables across multiple geographic resolutions (region and geo).
• Command-line interface for simplified usage and enhanced computational efficiency.
• Fully compatible with Python 3, with updated dependencies.

Run PopGen3 Online

• 
• 

Contents

Data Inputs

Survey sample data:

household_sample.csv person_sample.csv

Population marginal data:

• geo-level marginals:
  person_marginals.csv
household_marginals.csv

• region-level marginals:
  region_person_marginals.csv
region_household_marginals.csv

Multi-Geographic Resolution Mapping Data:

region_geo_mapping.csv
geo_sample_mapping.csv

Example of Multi-Geographic Resolution Levels:

• Region Level: Set as census county subdivision
• Geo Level: Set as census tracts
• Sample Geo Level: Set as Public Use Microdata Areas (PUMAs)

Configuration File

The configuration.yaml file contains several key sections for the PopGen setup.

Designed Input Keys

Project Wide Setting	Input Data Files	Scenario Settings
- synthesize - name - location	- entities - column_names - location	- description - apply_region_controls - control_variables - parameters - geos_to_synthesize - outputs

Key Outputs

Sample Weighting Results

• weights.csv

Synthetic Population Generated

• housing_synthetic.csv
• person_synthetic.csv