MARG - Mobility Analytics Research Group
  • Home
  • About
    • MARG Team
    • Research and Services
    • Collaborate and Contribute
  • Projects
    • PopGen
    • Vehicle Fleet Mix Model
    • R Codes for Travel Modeling
    • Integrated Model: SimTRAVEL
    • openAMOS
    • University Survey and Model
    • Tour Structure Model
    • Integrated Energy Model
    • Wellbeing Estimator
  • FOSS-TM
  • Contact
    • Guestbook

Synthetic Population Generator
PopGen

References
Software
PopGen 1.1 vs 2.0

Population Synthesis

Picture
The advent of activity-based microsimulation models of travel demand has provided the ability to simulate activity-travel patterns of individual travelers in time and space. The application of activity-based travel demand model systems requires the generation of synthetic populations whose attribute distributions match those of the general population within small geographies (say, a traffic analysis zone, census tract, block group, or block). The generation of synthetic populations has therefore received considerable attention in the profession and several synthetic population generators or population synthesizers have been developed and applied in various urban areas that are transitioning to or applying activity-based model systems. The Mobility Analytics Research Group developed a new heuristic algorithm for iteratively generating a representative synthetic population. In some older synthetic population generators, the distributions of household-level attributes match well with census distributions, but the distributions of person-level attributes are not adequately controlled to ensure the best possible match with census distributions. The new algorithm developed and incorporated in PopGen is capable of producing synthetic populations while controlling and matching both household-level and person-level attribute distributions. The PopGen software can be implemented easily and effectively for synthesizing populations while matching population controls in small geographies. PopGen has been applied and tested in a number of states, metropolitan planning organizations (MPOs), and research studies. PopGen is capable of synthesizing populations whose attribute distributions closely match those of the general population as documented in the census. The Mobility Analytics Research Group is working on the next generation of PopGen, which will incorporate full population evolution and socio-economic dynamics to evolve the population over time.
            The development of PopGen was made possible through generous support provided by the Federal Highway Administration under its Exploratory Advanced Research Program (EARP).  Additional support was provided by Arizona State University, Georgia Institute of Technology, and University of Connecticut through faculty start-up resources.  In addition, a number of agencies have supported and sponsored the development, refinement, and enhancement of PopGen.  In particular, Maricopa Association of Governments (MAG), Baltimore Metropolitan Council (BMC), Southern California Association of Governments (SCAG), and Colorado Department of Transportation (CDOT) have provided support for and implemented/tested PopGen.  

 

References

Top
The following publications offer details about the methods embedded in the PopGen software.  Please cite these articles in any work that utilizes the PopGen synthetic population generator software system. 
  1. Konduri, K.C., D. You, V.M. Garikapati, and R.M. Pendyala (2016) Application of an Enhanced Population Synthesis Model that Accommodates Controls at Multiple Geographic Resolutions. Transportation Research Record, Journal of the Transportation Research Board (forthcoming). 
  2. Ye, X., K. Konduri, R.M. Pendyala, B. Sana, and P. Waddell (2009) A Methodology to Match Distributions of Both Household and Person Attributes in the Generation of Synthetic Populations. Proceedings of 88th Annual Meeting of the Transportation Research Board,  National Research Council, Washington, D.C. 
  3. Bar-Gera, H., K. Konduri, B. Sana, X. Ye, and R.M. Pendyala (2009) Estimating Survey Weights with Multiple Constraints Using Entropy Optimization Methods. Proceedings of 88th Annual Meeting of the Transportation Research Board,  National Research Council, Washington, D.C.
In addition, please cite this webpage as follows:
  • MARG (2016) PopGen: Synthetic Population Generator [online]. Mobility Analytics Research Group. Available at: http://www.mobilityanalytics.org/popgen.html, Accessed [Date of Access]. 
There are a few legacy documents and websites that provide additional information about PopGen.  Please note that the information may be outdated and some weblinks embedded in these resources are likely to be non-functional.  They are: 
  1. Pendyala, R.M., K.C. Konduri, and K.P. Christian (2011) PopGen 1.1 User’s Manual.  Lulu.com Publishers, USA.  
  2. http://urbanmodel.asu.edu/popgen.html  
  3. http://simtravel.wikispaces.asu.edu/Population+Synthesis    
 

PopGen 1.1

Download and unzip (if applicable) all of the files below.  Then, follow the instructions for installation exactly as specified.  All files below may be downloaded to a single directory.  PopGen 1.1 works with all recent versions of the Windows operating system. 
Software
Release: Apr 20, 2016
Distributed Under:
License Agreement

Check Out:
PopGen 1.1 Github​

​PopGen 1.1 Presentation
Program Files
PopGen 1.1 (.zip)
File Size: 13929 kb
File Type: zip
Download File

Sample Input Data (.zip)
File Size: 95 kb
File Type: zip
Download File

User Documentation (.zip)
File Size: 10317 kb
File Type: zip
Download File

Supplementary Packages
mysql-5.1.73-winx64.msi
File Size: 104240 kb
File Type: msi
Download File

scipy-0.15.1-cp27-none-win_amd64.zip
File Size: 89790 kb
File Type: zip
Download File

Other Supporting Packages (.zip)
File Size: 86025 kb
File Type: zip
Download File


PopGen 2.0

Software
Release: April 20, 2016

Distributed Under:
License Agreement

Check Out:

PopGen 2.0 Github

PopGen 2.0 Presentation

Download and unzip (if applicable) all of the files below.  Then, follow the instructions for installation exactly as specified.  All files below may be downloaded to a single directory.  PopGen 2.0 works with all recent versions of the Windows operating system. 
Program Files
Popgen 2.0 (.zip)
File Size: 155 kb
File Type: zip
Download File

Supplementary Packages
scipy-0.15.1-cp27-none-win_amd64.zip
File Size: 89790 kb
File Type: zip
Download File

Sample Input Data (.zip)
File Size: 36 kb
File Type: zip
Download File

Other Supporting Packages_PopGen2 (.zip)
File Size: 52739 kb
File Type: zip
Download File

UserGuide.zip
File Size: 2905 kb
File Type: zip
Download File

Top
 

PopGen 1.1 vs PopGen 2.0

Top
Both PopGen 1.1 and PopGen 2.0 are fully functional software systems.  They are some key differences between the two versions of the software that users may consider before selecting the specific version to use for an application. 
PopGen 1.1
  • Accommodates household and person control variables at one level of geography, i.e., all controls must be available at a single (same) geographic level (e.g., traffic analysis zone)
  • Has a fully functional graphical user interface (GUI) that makes project setup very easy and intuitive through a step-by-step wizard
  • Incorporates the ability to automatically download and process census data sets if synthetic population generation is being done at the level of county, block group, or census tract
  • Incorporates a number of data manipulation features including the ability to adjust household size control distribution to account for inconsistencies between household-level and person-level controls
  • Generates graphical output displays to quickly compare and assess goodness-of-fit of the synthetic population, including fit against controlled and uncontrolled variables and their distributions
  • Incorporates both iterative proportional updating (IPU) and entropy-based methods for synthetic population generation
  • Includes three different rounding schemes - bucket rounding, arithmetic rounding, and stochastic rounding to determine round cell frequencies
PopGen 2.0
  • Incorporates the ability to account for control distributions available at multiple (different) geographic scales
  • ​Does not have a graphical user interface (GUI) and software is run through command line interface 
  • Incorporates latest software coding standards, bringing about substantial efficiency in computational run times (runs 3-4 times faster than PopGen 1.1)
  • All project setup and input data preparation must be done offline (outside PopGen 2.0) and then specified in the configuration file (yaml file)
  • Uses new configuration file setup (yaml) which is easy to use and very intuitive
  • Provides ability to output customized multi-way tabulations for variables of interest in the synthetic population
  • Incorporates both iterative proportional updating (IPU) and entropy-based methods for synthetic population generation
  • Includes only one rounding scheme - bucket rounding - to determine frequencies in joint distributions
  • Does not incorporate any automated census data downloading and processing tools
Top
Powered by Create your own unique website with customizable templates.
  • Home
  • About
    • MARG Team
    • Research and Services
    • Collaborate and Contribute
  • Projects
    • PopGen
    • Vehicle Fleet Mix Model
    • R Codes for Travel Modeling
    • Integrated Model: SimTRAVEL
    • openAMOS
    • University Survey and Model
    • Tour Structure Model
    • Integrated Energy Model
    • Wellbeing Estimator
  • FOSS-TM
  • Contact
    • Guestbook