DECOMP
Software for Multiple Standardization and Demographic
Decomposition
Steven Ruggles
ruggles@umn.edu
DECOMP is a program
designed to do multiple direct standardization and decomposition of differences
between two rates according to the method described by Prithwis Das Gupta,
"A General Method of Decomposing a Difference Between Two Rates into
Several Components," Demography
15:1 (1978), 99-111.
I wrote the program
gradually between 1986 and 1989, and I haven’t modified it since. I still use
it, and I know of several others who are still using it as of 2006. Feel free
to do anything you like with it. If you make any improvements, please send me a
copy. It would be terrific if somebody wanted to make a Windows version. If you
use it in a publication, please cite it appropriately and also let me know.
This program was
originally written using Microsoft Fortran for DOS, and I distributed it on
orange 5.25” 360K floppies. I have also included the source code for a UNIX F77
version (dc.f).
DECOMP reads
column-format ASCII data files using a simple set of SPSS-like commands. To run
the program in Windows, make a folder and copy the DOS executable (dc.exe) plus
the setup file (setup.cmd). Then select “run” from the start menu, and type
“cmd” in the dialog box to open a DOS window. Use DOS commands to navigate to
the directory that contains the program, and follow the instructions in the
manual.
|
DECOMP Manual (must
read!): |
|
|
Executable file: |
|
|
Setup file: |
|
|
|
|
|
Sample DECOMP command
files: |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Data for use with
sample command files: |
|
|
Codebook for sample
dataset: |
|
|
|
|
|
Source code (Sun F77
compiler) |
|
|
Source code (Microsoft
Fortran 4.1) |
|
|
|
|
|
Sample article using
DECOMP: |
|
|
|
|