[R] Numerical integration

nany23 anna.botto at gmail.com
Fri Jul 1 14:47:40 CEST 2011

thanks for the Italian! 
I apologize for my previuos explanation which was not clear
actually there are two "k" parameters, so I change one  them; let's put it
this way

/# these are the 3 parameters 
a<- 414.566 
b<- 345.5445 
g<- -0.9695679 
xstar<- 1397.923 
#I create a vector 
pars <-expand.grid(xstar = xstar, a= a, b= b , g= g) 
fun <- function(xstar, a,b,g,*m*) { 
    f <- function(x, xstar, a, b,g,m) f.lognorm(x) * *m* * x * (x >= xstar) 
    integrate(f, -Inf, Inf, xstar = xstar, a = a, b =b, g=g, *m=m*)$value 
# Method 1: (outputs a data frame) 
out <- mdply(pars, fun)  

at this stage a warning message comes out: Errore in k > -1e-07 : 'k' is

The "k" the errore refers to is one of the three parameters of the pdf
distribution whose formula is the following

/function (x, xi, alfa, k) 
    if ((*k > -1e-07*) & (k < 1e-07)) {
        y <- (x - xi)/alfa
    else {
        y <- -k^(-1) * log(1 - k * (x - xi)/alfa)
    f <- exp(k * y - (y^2)/2)/(alfa * sqrt(2 * pi))

So the xi,alfa, k of the function are those which I call a,b,g[the
I'm afraid I'm making some very silly mistakes in the syntax but don't know
where and how correct them... 

I tried different ways but they don' work...

View this message in context: http://r.789695.n4.nabble.com/Numerical-integration-tp3634365p3638104.html
Sent from the R help mailing list archive at Nabble.com.

More information about the R-help mailing list