R/SummaryOptim.R

In jinseob2kim/MDLM: Multi Dimensional Linear Model(MDLM)

###################################################################################################
# Function: summaryoptim
#
# Make summary table of model fit from optim(): Estimate, Std.Error, z.value, p.value.
#
#
# ARGUMENTS:
#
# optim.fit: Fitted model object from optim().
#
# VALUE:
#
# ds.out: Table with 4 columns: cbind(Estimate, Std.Error, z.value, p.value).
#
# Author: DVN
# Senturk D, Dalrymple DS, Mu Y, Nguyen DV (2014) Weighted hurdle regression method for joint
# modeling of cardiovascular events likelihood and rate in the U.S. dialysis population.
#
# 2.8.14
###################################################################################################



summaryoptim <- function ( optim.fit, N=100 ){

  Estimate = optim.fit$par
  #cov.mat = solve(optim.fit$hessian)
  cov.mat= 2*optim.fit$value/(N - length(Estimate)) * solve(optim.fit$hessian)
  Std.Error = sqrt( diag(cov.mat)	)
  z.value = Estimate/Std.Error
  p.value = 2*( 1-pnorm(abs(z.value)) )

  ds.out = cbind(Estimate, Std.Error, z.value, p.value)
  colnames(ds.out) <- c('Estimate', 'Std Error', 'z value', 'p value')

  ds.out
}
###################################################################################################


#library(numDeriv)
#' @import numDeriv
#' @importFrom stats optim pnorm
summaryoptim2 <- function ( optim.fit,f, N=100 ){

  Estimate = optim.fit$par
  #cov.mat = solve(optim.fit$hessian)
  cov.mat= 2*optim.fit$value/(N - length(Estimate)) * solve(hessian(f,optim.fit$par))
  Std.Error = sqrt( diag(cov.mat)	)
  z.value = Estimate/Std.Error
  p.value = 2*( 1-pnorm(abs(z.value)) )

  ds.out = cbind(Estimate, Std.Error, z.value, p.value)
  colnames(ds.out) <- c('Estimate', 'Std Error', 'z value', 'p value')

  ds.out
}

#' @importFrom optimx optimx
summaryoptim3 <- function ( optimx.fit, f, N=100 ){
  name_value = which(names(optimx.fit)=="value")
  Estimate = as.numeric(optimx.fit[1:(name_value-1)])
  #cov.mat = solve(optim.fit$hessian)
  cov.mat= 2*as.numeric(optimx.fit[name_value])/(N - length(Estimate)) * solve(hessian(f,Estimate))
  Std.Error = sqrt( diag(cov.mat)	)
  z.value = Estimate/Std.Error
  p.value = 2*( 1-pnorm(abs(z.value)) )

  ds.out = cbind(Estimate, Std.Error, z.value, p.value)
  colnames(ds.out) <- c('Estimate', 'Std Error', 'z value', 'p value')

  ds.out
}