Fitting FMKL GLD

Share:

Description

To fit a FMKL GLD to raw/binned data.

Usage

1
2
 fit.GLD.FMKL(x, lbound, ubound, percentile='exact', mle=FALSE)
 fit.GLD(x, lbound, ubound, method='chisquare')

Arguments

x

A vector of raw data, or a histogram or binned data.

percentile

Use the exact percentiles (exact) or approxiated values (approximate).

mle

Logical. To find the MLE or not.

lbound,ubound

lower and upper bound for the support of the density. The bounds could be finite values, or positive or negative infinity.

method

Method for goodness-of-fit test.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
  data(hhi)
  hmob <- binning(counts=hhi$mob, breaks=hhi$breaks)
  lmd5 <- fit.GLD.FMKL(hmob)
  lmd6 <-  fit.GLD.FMKL(hmob, mle=TRUE)
  plot(lmd5)
  lines(lmd6, col=4)
  ## GOP example (handbook) -- Hahn & Sapiro (1967)
  ## KS-GLD based on original data: (0.0345, 0.00009604, 0.87, 4.92)
  ## Table 3.6-1
  breaks <- c(-Inf, seq(0.015, length=10, by=0.005), Inf)
  counts <- c(1,9,30,44,58,45,29,17,9,4,4)
  rho.mid <- c(0.0325, 0.0250, 0.667, 0.600)
  rho.unif <- c(0.03352, 0.02531, 0.7786, 0.5009)
  ## histogram for chi-square test
  ## KS = 0.0225, p-value = 0.999.  Chi=0.5176, p-value=0.7720
  breaks <- c(-Inf, 0.025, 0.03, 0.035, 0.04, 0.045, 0.05, Inf)
  counts <- c(40,44,58,45, 29,17,17)