Outlier Detection Tools for Functional Data Analysis

A collection of functions for outlier detection in functional data analysis. Methods implemented include directional outlyingness by Dai and Genton (2019) , MS-plot by Dai and Genton (2018) , total variation depth and modified shape similarity index by Huang and Sun (2019) , and sequential transformations by Dai et al. (2020)

fdaoutlier

Outlier Detection Tools for Functional Data Analysis

tools for functional data analysis. Methods implemented include directional outlyingness, MS-plot, total variation depth, and sequential transformations among others.

Installation

You can install the current version of fdaoutliers from CRAN with:

install.packages("fdaoutlier")

or the latest the development version from GitHub with:

devtools::install_github("otsegun/fdaoutlier")

Example

Generate some functional data with magnitude outliers:

library(fdaoutlier)
simdata <- simulation_model1(plot = T, seed = 1)

dim(simdata$data)
#> [1] 100  50

Next apply the msplot of Dai & Genton (2018)

ms <- msplot(simdata$data)

ms$outliers
#> [1]  4  7 17 26 29 55 62 66 76
simdata$true_outliers
#> [1]  4  7 17 55 66

Methods Implemented

1. MS-Plot (Dai & Genton, 2018)
2. TVDMSS (Huang & Sun, 2019)
3. Extremal depth (Narisetty & Nair, 2016)
4. Extreme rank length depth (Myllymäki et al., 2017; Dai et al., 2020)
5. Directional quantile (Myllymäki et al., 2017; Dai et al., 2020)
6. Fast band depth and modified band depth (Sun et al., 2012)
7. Directional Outlyingness (Dai & Genton, 2019)
8. Sequential transformation (Dai et al., 2020)

Bugs and Feature Requests

Kindly open an issue using Github issues.