cg2

CG2 CG2 Fixed Effects Estimation Software =====================================

For more details, see http://labordynamicsinstitute.github.io/cg2 (latest).

Overview

CG2 is a package of Fortran 90 and SAS programs that estimate non-nested 2 component fixed effect models. The programs have been successfully compiled and tested on multiple Unix platforms (IA32 (SuSe and Redhat Linux), Itanium (SuSe Linux), and Sparc (Solaris)), but the code should work fine on other platforms as well (ie. IA32 (Windows)). A port to Stata (a2reg) is maintained by Amine Ouzad (an early version is available in the Stata branch, the latest version can be found at https://ideas.repec.org/c/boc/bocode/s456942.html).

The estimation algorithms were developed to solve large-scale fixed person and firm effect wage models. In the typical scenario, a person’s earnings and place of employment are observed over time, with mobility of persons across firms. This mobility or “connectedness” of persons enables the estimation of the model, although for problems with millions of persons and firms, obtaining results require a substantial amount of computing power. For users with large problems to solve, the main constraint will be acquiring a computing platform with sufficient physical memory or RAM. For example, the largest problem we have solved required 137GB of physical memory and took zzzz hours of CPU time on a Sun Fire 12k server. However, this does not imply that CG2 is inefficient, since no known software can solve the same size problem using fewer system resources. To further reduce single system image memory requirements, a cluster aware version of CG2 is under development.

The description here was first created several years ago. References to obsolete hardware (IA32, Sun Fire 12k server ) are to be excused.

If you need additional information, please see the [References][#References] section below.

Requirements

In order to use the CG2 package, certain basic software must be available on a current Unix platform (The software will likely port to Windows just fine, but we have not devoted time to testing this assertion. If anyone succeeds in using this software on IA32 Windows, please let us know and we will post the information here).

Even if the above requirements can be met, the user must insure that their computing platform has sufficient physical memory. The calculator will allow such an estimation.

Download

Download the distribution using the buttons on this page, or clone the repository.

Installation

After download, there should be three directories:

Using CG2

Before actually estimating model parameters, you must properly structure the input file and calculate the size of the problem. With the input data ready, the next step is to run CG2 and calculate parameter estimates. The final step is post-processing the estimates, although this may not be required depending on the set of parameters you are interested in.

Test data

The SAS programs as packaged are setup to use a purely fictional, synthetic dataset. It is a good idea to run through each HowTo on the synthetic data first before attempting a problem of your own. This will give you an opportunity to become familiar with each stage of the process as well as test your installation.

Typographical convention

We have tried to be consistent throughout in the use of typographical convention:

Howtos

Acknowledgements

John Abowd @johnmabowd, Robert Creecy, Kevin McKinney, and Lars Vilhuber @larsvilhuber. Census Bureau and the rest of the LEHD staff.

Contact

If you have any questions or comments please contact Kevin McKinney at kevinm@ccrdc.ucla.edu.

References