R/est_mfa.R

est.mfa <- function (init_para, Y, itmax, tol, conv_measure, ...) {

p <- ncol(Y)
n <- nrow(Y)
fit <- init_para

loglikeNtau <- try(do.call('logL_tau.mfa', c(list(Y = Y), fit)),
                        silent = TRUE)

if ((any(class(loglikeNtau) %in% "try-error")) ||
              (any(class(loglikeNtau) %in% 'character'))) {
  FIT <- paste('in computing the log-likelihood before EM-steps')
  class(FIT) <- "error"
  return(FIT)
}

fit <- append(fit, loglikeNtau)

for (niter in 1 : itmax) {

  FIT <- do.call('Mstep.mfa', c(list(Y = Y), fit))

  if (any(class(FIT) %in% 'error')) {
    FIT <- paste('in ', niter,
                   'iteration of the M-step', FIT)
    class(FIT) <- "error"
    return(FIT)
  }

  loglikeNtau <- try(do.call('logL_tau.mfa', c(list(Y = Y), FIT)),
                                     silent = TRUE)

  if ((any(class(loglikeNtau) %in% "try-error")) ||
              (any(class(loglikeNtau) %in% 'character'))) {
    FIT <- paste('in computing the log-likelihood after the ', niter,
                   'th the M-step', FIT$logL, sep = '')
    class(FIT) <- "error"
    return(FIT)
  }

  FIT <- append(FIT, loglikeNtau)
  
  if ((any(class(FIT$logL) == "NULL")) || 
      (any(class(FIT$logL) == 'character'))) {

    FIT <- paste('in computing the log-likelihood after the ', niter,
                     'th the M-step', FIT$logL, sep = '')
    class(FIT) <- "error"
    return(FIT)
  } else {
    if ((FIT$logL == -Inf) | is.na(FIT$logL)) {
      FIT <- paste('the log-likelihood computed after the ', niter,
                   'th iteration of the M-step is not finite', sep = '')
      class(FIT) <- "error"
      return(FIT)
    }
  }

  if ((conv_measure == "diff") && (abs(FIT$logL - fit$logL) < tol))
    break
  if ((conv_measure == "ratio") && (abs((FIT$logL - fit$logL) / FIT$logL) < tol))
    break

  fit <- FIT
}

class(FIT) <- "mfa"
return(FIT)
}

Try the EMMIXmfa package in your browser

Any scripts or data that you put into this service are public.

EMMIXmfa documentation built on Dec. 18, 2019, 1:40 a.m.