bbemkr (version 1.2)

bbewarmup: Burn-in period

Description

By minimizing the cost value, the function estimates the bandwidths of the regressors and error variance parameter of the error density for the warmup or burn-in period

Usage

bbewarmup(x, costpara, kerntype = c("Gaussian", "Epanechnikov", 
          "Quartic", "Triweight", "Triangular", "Uniform"), 
          warm = 4, sizep = 2)

Arguments

x

Log bandwidths of the regressors

costpara

Initial cost value

kerntype

Type of kernel function. By default, Gaussian kernel is used

warm

Number of iterations in the warmup or burn-in period

sizep

A tuning parameter used in the random-walk Metropolis algorithm

Value

• xh
  Log bandwidths of the regressors, obtained after the warmup period
• xhtest
  Log bandwidths of the regressors, obtained during the warmup period
• sigma
  Variance of the error density, obtained after the warmup period
• sigmatest
  Variance of the error density, obtained during the warmup period
• cost
  Cost value, obtained after the warmup period
• costtest
  Cost value, obtained during the warmup period
• accept_rate
  Acceptance rate of the random-walk Metropolis algorithm

See Also

bbeMCMCrecording, bbelogdensity