BIC_tuner: #' @keywords internal qdet <- function(local_q = 0.1, nz)...

In hoangtt1989/NOIS: Non-parametric Outlier Identification and Smoothing

Description

Find the optimal number of detected outliers for NOIS using modified BIC

Usage

BIC_tuner(data, q_tst = 1:floor(nrow(data)/3), bias_correct = T,
  parallel = F, return_fit = T, ...)

Arguments

data	An input data.frame.
q_tst	The grid of test values for the pooled q.
bias_correct	A logical indicating bias correction.
parallel	A logical indicating parallel computation. A parallel backend must be registered.
return_fit	A logical indicating whether a model with the optimal q should be fit (using NOIS_fit) and returned.
...	Additional arguments passed to NOIS_fit.

Value

A list with the following components.

q_tst	The grid of test values.
min_q	The optimal number of detected outliers.
opt_fit	A NOIS_fit using the optimal number of detected outliers.
BIC_vals	The BIC for each fit.
df_vals	The estimated degrees of freedom for each fit.

See Also

Other NOIS BIC functions.: BIC_NOIS

hoangtt1989/NOIS documentation built on May 20, 2019, 2:08 p.m.