Copyright 1999-2000 VA Linux Systems, Inc.LuaForge Project News
http://luaforge.net
LuaForge Project News Highlightsinequality-2.0.1
http://luaforge.net/forum/forum.php?forum_id=1261
=== 2.0.1 ===
- Python script added to Lua script (core of
http://www.poorcity.richcity.org/cgi-bin/inequality.cgi)
- Display of Median improved
- Coulter not displayed anymore
- Avoidance of numeric computation errors on some platforms
- inequalities.html (small calculus) added to zip archiveInequality MeasuresGoetz KlugeMon, 16 Jul 2007 2:06:08 GMTinequality-2.0.0
http://luaforge.net/forum/forum.php?forum_id=1260
In the new Lua-script the display of redundancies
("Theil" and "symmetric") has been reactivated and
the computation of averages for recources (e.g.
income) has been added: Mean, median and welfare
functions.Inequality MeasuresGoetz KlugeTue, 10 Jul 2007 4:16:23 GMTinequality-1.4.0
http://luaforge.net/forum/forum.php?forum_id=1250
inequality-1.4.0.lua and inequality-1.4.0.py are an update of inequality-1.3.4 (from 2004). I also added a brief collection of all used formulas to the zipped packet (inequality.html).
The scritps compute inequality measures (Gini, Atkinson, Theil etc.) and is an application for social sciences (econometrics, welfare economics etc.)
Changes:
(1) Optimized for Lua 5.1
(2) Coulter inequality corrected
(3) "Kullback-Leibler Inequality" renamed to "Symmetric Inequality"Inequality MeasuresGoetz KlugeFri, 06 Jul 2007 10:31:50 GMTInequality Measures: Flat-Tax and Income Distribution
http://luaforge.net/forum/forum.php?forum_id=631
FTcalc-2001-1.0.1 corrects error in FT calculation in FTcalc-2001-1.0.0.
By the way:
(a) There is an experimental Syntropy tax computation.
(b) Sorry, the text mostly is in German.
Inequality MeasuresGoetz KlugeSat, 03 Dec 2005 11:21:30 GMTinequality-task-1.0.1
http://luaforge.net/forum/forum.php?forum_id=466
inequality-task-1.0.1.lua is inequality-1.3.4.lua with file input. This is a spin off from a customized version for a social research project in South America. There is a small package (with script for all platforms, a sample input file and Lua 5.0.2 for MS-Windows) and a big package (like small package, plus lots of docu on Lua and a SciTE 1.6.2 editor).Inequality MeasuresGoetz KlugeSat, 09 Apr 2005 9:54:56 GMTinequality-1.3.4 sources
http://luaforge.net/forum/forum.php?forum_id=463
(1) The plain source code now is available too. (Previously only fancy colored HTML pages were avaible from which source code could be extracted when saving the pares as text files.)
(2) http://www.poorcity.richcity.org/calculator.htm is a on-line calculator using investor-1.3.4Inequality MeasuresGoetz KlugeWed, 06 Apr 2005 2:01:20 GMTinequality.cgi V1.3.3
http://luaforge.net/forum/forum.php?forum_id=261
inequality-1.3.3 has been implemented as CGI application:
http://www.poorcity.richcity.org/calculator.htmInequality MeasuresGoetz KlugeTue, 14 Sep 2004 8:03:08 GMTinequality.cgi V1.1.3
http://luaforge.net/forum/forum.php?forum_id=260
inequality-1.3.3 has been implemented for CGI:
http://www.poorcity.richcity.org/calculator.htm
Inequality MeasuresGoetz KlugeTue, 14 Sep 2004 7:31:51 GMTinequality-1.3.3
http://luaforge.net/forum/forum.php?forum_id=209
(1) The program has been used to generate histograms which show the frequency of inequality indexes when feeding random data to the program. I did that in Python and also in Lua.
(2) The results of the experiments are part of inequality-1.3.3.zip.
(3) The random data contained some weird constellations which never will show up in real life. But these "impossible" data helped me to improve the exactnes of the computation of the Platon index.
(4) The Platon index is a transformation of the Kullback-Leibler index. Interestingly, it behaved very much like the Gini index in the random data tests.
(5) I deleted the news on 1.3.0 in order not to clutter this board with news.
(With 1.3.0 the Williamson variation had been fixed. It can be used as population weighted standard variation. But I don't recommend it. Rather, I recommend to work with relative inequalities, especially with the Kullback-Leibler inequality and the Platon inequality. Scaled between 0 and 1 (o% and 100%), relative inequalities are the most "readible" indicators for inequality. And they are suitable to compute Amartya San's welfare function, like e.g. GNP*(1-relative_inequality).
(6) What is the difference between redundancies and inequalities? Redundancies are the difference between max. entropy and a iven entropy of a system. Redundancies R (e.g. the Kullback-Leibler-Distance) can be converted into relative inequalities Z: Z=1-exp(-R). Inequality-1.3.3 now only displays inequalities, but also could display some redundancies (Theil, Atkinson and Kullback-Leibler) after unlocking a few lines in the source code.Inequality MeasuresGoetz KlugeSun, 01 Aug 2004 9:00:50 GMTInequality 1.3.0
http://luaforge.net/forum/forum.php?forum_id=201
The Williamson variation has been fixed. It can be used as population weighted standard variation.
But I recommend to work with relative inequalities, especially with the Kullback-Leibler inequality. Scaled between 0 and 1 (o% and 100%), relative inequalities are the most "readible" indicators for inequality. And they are suitable to compute Amartya San's welfare function, like e.g. GNP*(1-relative_inequality). Redundancies R (e.g. the Kullback-Leibler-Distance) can be converted into relative inequalities Z: Z=1-exp(-R).Inequality MeasuresGoetz KlugeSun, 01 Aug 2004 4:12:00 GMT