Dear fellows !
I made numerical convolution with Stata in order to verify that
dosis-distribution
in human population is composed of two lognormal curves.
I want to emphasize that this is only verification.
-
Now I have an idea of deconvolution to be mathematical 'stronger ':
I have 200 integral-equations of the form :
/please excuse pseudocode /
- a1 = integ(0,k1)[exp(-(((k1-x)-m1)^2/(2*s1^2))) *
*exp(-((x-m2)^2/(2*s2^2))) ]
- a2 = integ(0,k2)[exp(-(((k2-x)-m1)^2/(2*s1^2))) *
*exp(-((x-m2)^2/(2*s2^2))) ]
.
.
.
.
.
- an = integ(0,kn)[exp(-(((kn-x)-m1)^2/(2*s1^2))) *
*exp(-((x-m2)^2/(2*s2^2))) ]
n: 1.... 200 the couples (ai,ki),i = 1..200are well known
I want to know (m1,s1,m2,s2 )
I'd like to solve this system with Stata,which is the most quickiest way ?
is it possible ?
Why with Stata?because our aim is to get data from somewhere in Europe
then we want to haul out a subgroup ,show how to verify convolution-
as I already did - and at last solve equation-system above to get parameters
for the both components:background and add on.
If I know components I can always compute add on radioactivity which some
personal group is exposed - this would be new - and I would get nobel price.
(this was a joke)
sincerely aschbacher andreas
--
+++ NEU bei GMX und erstmalig in Deutschland: T�V-gepr�fter Virenschutz +++
100% Virenerkennung nach Wildlist. Infos: http://www.gmx.net/virenschutz
*
* 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/
