Outliers due to technical errors in water-quality data from in situ sensors can reduce data quality and have a direct impact on inference drawn from subsequent data analysis. Here, we proposed an automated framework that provides early detection of outliers in water-quality data from in situ sensors caused by technical issues. The framework was used first to identify the data features that differentiate outlying instances from typical behaviours. Then statistical transformations were applied to make the outlying instances stand out in transformed data space. Unsupervised outlier scoring techniques were then applied to the transformed data space and an approach based on extreme value theory was used to calculate a threshold for each potential outlier.

In this work we showed that our proposed framework, with carefully selected data transformation methods derived from data features, can greatly assist in increasing the performance of a range of existing outlier detection algorithms.

This package is still under development and this repository contains a development version of the R package oddwater.

You can install oddwater from github with:

# install.packages("devtools")
devtools::install_github("pridiltal/oddwater")

library(tidyverse)
library(oddwater)
data("data_sandy_anom")
head(data_sandy_anom)
#>         Timestamp Level   Cond   Tur label_Level label_Cond label_Tur
#> 1 12/03/2017 1:00 0.636 326.34 34.47           0          0         0
#> 2 12/03/2017 2:30 0.636 326.63 34.06           0          0         0
#> 3 12/03/2017 4:00 0.635 327.00 33.39           0          0         0
#> 4 12/03/2017 5:30 0.633 327.31 32.61           0          0         0
#> 5 12/03/2017 7:00 0.634 327.81 33.34           0          0         0
#> 6 12/03/2017 8:30 0.631 328.70 32.22           0          0         0
#>   type_Level type_Cond type_Tur
#> 1          0         0        0
#> 2          0         0        0
#> 3          0         0        0
#> 4          0         0        0
#> 5          0         0        0
#> 6          0         0        0

oddwater::explore_data()