UnivOutl
The contents related to univOutl are shown in the following sections:
The univOutl package is package for the R environment that implements the main techniques for identifying outliers in data related to a single quantitative variable (univariate outliers) or in cases where the same quantitative variable is observed on the same units in different occasions, such as in panel surveys.
The methods for identifying univariate outliers are essentially based on two approaches:
- The assumption of a Gaussian model for the data distribution;
- A non-parametric approach based on the use of boxplots.
When a Gaussian distribution is assumed, the univOutl package offers various solutions for the robust estimation of distribution parameters (mean and standard deviation) to reduce the influence of extreme values.
The non-parametric approach, on the other hand, is more flexible as it does not require a specific distributional model; it only requires an evaluation of the degree of symmetry in the data distribution. With symmetric distributions the outliers are identified using the traditional boxplot. In presence of asymmetric distributions, a common situation for variables related to businesses, farms, or household economic data, methods based on boxplots are appropriately modified to account for skewness.
Finally, the package includes specific tools for identifying outliers in continuous variables observed on the same units at two different points in time. These methods are based on the construction of ratios, including the Hidiroglou-Berthelot (1986) method and a non-parametric extension of it which introduces an additional degree of flexibility.
It should also be noted that, within the context of sample surveys, several functions in the univOutl package allow for the incorporation of survey weights into the analysis.
Main references
Hidiroglou, M.A. and Berthelot, J.-M. (1986) “Statistical editing and Imputation for Periodic Business Surveys”. Survey Methodology, Vol 12, pp. 73-83.
McGill, R., Tukey, J. W. and Larsen, W. A. (1978) “Variations of box plots”. The American Statistician, 32, pp. 12-16.
Rousseeuw, P.J. and Croux, C. (1993) “Alternatives to the Median Absolute Deviation”, Journal of the American Statistical Association, 88, pp. 1273-1283.
Hubert, M., and Vandervieren, E. (2008) “An Adjusted Boxplot for Skewed Distributions”, Computational Statistics & Data Analysis, 52, pp. 5186-5201
Status: validated
Author: Istat
Licence: GPL-2 | GPL-3
GSBPM code:5.4 Edit and impute
Programming language: R
Keywords: robust estimation; boxplot; Hidiroglou-Berthelot method
Contact: name: Marcello D’Orazio – email: madorazi@istat.it
TECHNICAL REQUIREMENTS
The univOutl package works on R versions 3.6.0 and later on any operating system (Windows, Mac, or Linux). It requires the following additional R packages to be installed and subsequently loaded: robustbase and Hmisc.
COPYRIGHT
Copyright 2026 Marcello D’Orazio
Licensed under the GNU General Public License (GPL), version 2 or later. You may not use this work except in compliance with the License. You may obtain a copy of the License at the following address: http://www.gnu.org/licenses/.
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
DISCLAIMER
Istat does not assume responsibility for results deriving from a use of the tool that is not consistent with the methodological indications contained in the available documentation.
DOWNLOAD
Release date: 21/02/2026
univOutl Version 0.5.0 – Precompiled package for Windows
univOutl Version 0.5.0 – Package source for Windows and Unix-like systems
INSTALLATION
In R the package can be installed using the following instructions:
> install.packages(path_to_file, repos = NULL)
where path_to_file indicated the path of the downloaded.zip or .tar.gz.
TECHNICAL AND METHODOLOGICAL DOCUMENTATION
Reference manual – univOutl v. 0.5.0
https://cran.r-project.org/web/packages/univOutl/univOutl.pdf
Hidiroglou, M.A. and Berthelot, J.-M. (1986) “Statistical editing and Imputation for Periodic Business Surveys”. Survey Methodology, Vol 12, pp. 73-83.
McGill, R., Tukey, J. W. and Larsen, W. A. (1978) “Variations of box plots”. The American Statistician, 32, pp. 12-16.
Rousseeuw, P.J. and Croux, C. (1993) “Alternatives to the Median Absolute Deviation”, Journal of the American Statistical Association, 88, pp. 1273-1283.
Hubert, M., and Vandervieren, E. (2008) “An Adjusted Boxplot for Skewed Distributions”, Computational Statistics & Data Analysis, 52, pp. 5186-5201
OTHER DOCUMENTATION
https://github.com/marcellodo/univOutl