I am trying to fit an appropriate distribution (perhaps a lognormal or
gamma or singh-madalla,...) to an income distribution (of counts) from the
US census. Basicly I want to get the quartiles of the income distribution
for each census tract of my study area. For example, census tract #1(a
row in your dataset) would have a count of the number of people in that
census tract with annual income of <$10,000 (a column), $10,000-$14,999
(another column), $15,000-$19,999 (another column), ect ect. I have seen
a couple of programs that fit a distribution using micro data, but I have
not seen a program that fits a distribution using grouped/banded data like
the example above. It seems to me that this would be a fairly common
problem associated with distributions from the census. Does anyone know
of a program in stata that would do this so I can ultimately get the
quartiles from the fitted distribution? I am guessing the program would
use the ML estimator in stata, but perhaps there is some type of kernal
estimation routine.
Thanks in advance,
Jaren Pope
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
