An observation.
- From: Les Cargill <lcargill@xxxxxxxxxx>
- Date: Fri, 19 Oct 2007 18:04:40 -0400
y = 34.856 Ln(x) + 1.2837
where x is roughly (income/$10,000) [which is how the chart is built].
That's the curve Excel shows as a remarkably good ( sorry, no Kolmogrov for now, if that even works on ln() based curves - I forget ) fit for
a household income CDF, based on the data at:
http://en.wikipedia.org/wiki/Image:Income-curve-%2410k.png
as part of
http://en.wikipedia.org/wiki/Household_income_in_the_United_States
Am I wrong? Clearly, ln() curves have no natural asymptope, so
it'd peter out based on some limit, preserving convergence
to 100%*, but it sure *LOOKS* good :) I'd expetc wierdness at
the tails...
*yes, that's the sound of one hand waving.... on the other hand...
--
Les Cargill
.
- Prev by Date: Re: Fear not, a consortium of bankers is going to rescue the market
- Next by Date: Re: engaging in working dialog with Georgists?
- Previous by thread: BANK FOR SALE IN TURKEY
- Next by thread: for your own health and wealth
- Index(es):
Relevant Pages
|
|
