gLog.ep: Adjust for zero values and compute log(abs(e)^p)

In AutoSEARCH: General-to-Specific (GETS) Modelling

Description

Adjusts a series called e - typically a series of residuals or financial returns - for zero values, so that the logarithm can be applied on the absolute pth. exponentiated values. Next, log(abs(e)^p) is computed

Usage

gLog.ep(e, zero.adj=0.1, p=2, na.replace=NA)

Arguments

e	numeric vector, time series or zoo object
zero.adj	numeric value between 0 and 1 (the quantile adjustment for zero values). The default 0.1 means zeros are replaced by the 10 percent quantile of abs(e) before taking the logarithm
p	numeric value greater than zero. The power of the log-volatility specification
na.replace	the value to replace NA values with. Default: na.replace=NA

Value

log(abs(e)^p), a numeric, where the zeros in e have been adjusted

Author(s)

Genaro Sucarrat, http://www.sucarrat.net/

AutoSEARCH documentation built on May 1, 2019, 10:31 p.m.