In zhicongz/AnomDetct: Anomaly Detection via Scan Statistics

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

AnomDetct

The goal of AnomDetct is to detect embedded clusters in a given continuous distributed sample. The cluster can be either a fix point cluster or a cluster with small variance. The underlying distribution, cluster locations and cluster size are not necessary to be specified.

Installation

You can install the released version of AnomDetct from CRAN with:

install.packages("AnomDetct")

And the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("zhicongz/AnomDetct")

Example

This is a basic example which shows you how to solve a common problem:

library(AnomDetct)
## generate sampled data with embedded cluster
set.seed(100);x <- c(rgamma(10000,2,0.05),rnorm(200,50,1))
res <- ultimate_detct(x, HRR_kernel = "gaussian", est_fun = "sbsp",
                      n_hz_sample = 50, n_hz_size = 200, MLE_unit = 5,
                      x_unit = 0.001)

## Returns
# 1. Quantile of clusters.
# 2. Location of clusters with differen significant level.
# 3. Plot

res