FactorCopulaModel (version 0.1.1)

ml1factor_f90: min negative log-likelihood for 1-factor copula with nlm()

Description

min negative log-likelihood for 1-factor copula with nlm()

Usage

ml1factor_f90(nq,start,udata,copname,LB=0,UB=40,ihess=FALSE,prlevel=0,
  mxiter=100,nu=3)

Value

MLE as nlm object (estimate, Hessian, SEs, nllk)

Arguments

nq

number of quadrature points

start

starting point (d-vector or m*d vector, e.g. 2*d vector for BB1)

udata

nxd matrix of uniform scores

copname

name of copula family such as "gumbel", "frank", "bb1", "t" (copname common for all variables)

LB

lower bound on parameters (scalar or same dimension as start)

UB

upper bound on parameters (scalar or same dimension as start)

ihess

flag for hessian option in nlm()

prlevel

printlevel for nlm()

mxiter

max number of iterations for nlm()

nu

degree of freedom parameter if copname ="t"

Examples

# See example in r1factor()