Hi,
I have some problems – I have a Cauchy distribution with density function
f(x) = sigma / (pi * (sigma^2 + (x- miu)^2) ),
where sigma = scale and miu = location (in my case sigma = 3, miu = 0), and I
have to find with bootstrap
E | sigma_estimated^3 – sigma^3 | (#),
where sigma_estimated, I guess, is bootstrap estimate.
So, I can find sigma with R function fitdistr() from the real data, witch are
generated with rcauchy(1000, 0, 3), I guess sigma_estimated I can find with
function fitdistr too, but I don’t know how to take just scale column. Anyone
knows?
Ok, if my bootstrap will have R=1000 I have to find (#) something like this:
E | sigma_estimated^3 – sigma^3 | = (1/R) * sum (r = 1:R) [|sigma_boot(r)^3 –
sigma^3|]?
Next problem – I have with kernel estimate a density f() function? How can I
do this with R?
And the last – with bootstrap I have to estimate a characteristic function,
witch is
c = exp{i*t*miu – sigma|t|} (in my case it will be c = exp{– sigma|t|})? Is it
possible to solve this problem with R?
Sorry for my broken English and huge thanks to everyone who will help me! :)
__________________________________________________
[[alternative HTML version deleted]]