clean_extremes_sigma: Clean extremes by normal sigma method

In chriszheng2016/zstmodelr: Stock Factor Models for China Stock Market Investment

Description

Identify extremes in data vector by normal sigma method, and replace extremes by NA or max/min limitation

Usage

clean_extremes_sigma(x, n_sigma = 3, extreme_value = c("limit", "NA"))

Arguments

x	a vector of data.
n_sigma	numbers of sigma to identify extremes, by default 3.
extreme_value	value to replace extremes, i.e. "limit","NA", by default "limit".

Details

Basing on assumption of data as normal distribution, i.e., X~N(μ,σ^2)

P(|X-μ|) > k*σ = \begin{cases} 0.317 & k=1 \\ 0.046 & k=2 \\ 0.03 & k=3 \\ \end{cases}

So we could take data beyond |3 sigma| as extremes.

Value

 a vector of data without extremes

chriszheng2016/zstmodelr documentation built on June 13, 2021, 8:59 p.m.