Dealing with Zeros and Missing Values in Compositional Data Sets Using Nonparametric Imputation

By Vera Pawlowsky-Glahn. 2003

601 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Environmental Engineering, Mechanics of Materials

View Paper: https://www.semanticscholar.org/paper/385faec75e1e31e2fa2c35919c399dd2ec46cbaa

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Environmental Engineering

In Mechanics of Materials

Citations in the corpus

Outlier Detection for Compositional Data Using Robust Methods (2008, methodology): “For all logratio transformations, the problem of values xi = 0 is solvable in many ways, e.g. Aitchison (1986, p. 268), and Mart́ın-Fernández, Barceló-Vidal, and Pawlowsky-Glahn (2003).”
Projection Pursuit Multivariate Transform (2014, background): “Where unequal sampling is a significant issue, data imputation methods may be considered to include additional observations (Little and Rubin 2002; Martin-Fernandez et al. 2003).”
On the interpretation of differences between groups for compositional data (2015, background): “Following Mateu-Figueras et al. (2013), this assumption is equivalent to assuming logratio normality and homocedasticity for the compositional term εk.”
Dealing with Zeros (2011, background): “Both from theoretical and practical viewpoints, multiplicative replacement shows better behavior (Mart́ın-Fernández et al., 2003) than other previously-proposed imputation techniques: additive replacement and simple substitution. The interested reader may refer to Mart́ın-Fernández and Thió-Henestro”
Spatial Regression Modeling for Compositional Data With Many Zeros (2013, methodology): “01, where a zero occurs, and rescale the remaining non-zero components as in the multiplicative strategy of Mart́ın-Fernández et al. (2003). This imputed dataset can now be used to fit the alr model, though we use the original, non-imputed dataset for model comparison via CRPS.”
NetCoMi: network construction and comparison for microbiome data in R (2020, neutral): “Several zero replacement strategies have been proposed [40–42, 44, 46].”
Imputation of missing values for compositional data using classical and robust methods (2010, methodology): “This is also true for imputation methods (Bren et al., 2008; Martín-Fernández et al., 2003; Van den Boogaart et al., 2006; Palarea-Albaladejo and Martín-Fernández, 2008).”
Reply to “On the Harker Variation Diagrams; …” by J.A. Cortés (2009, background): “Based on a number of contributions (Aitchison and Egozcue 2005; Aitchison 1986), as well as some additional key contributions (Barcelo-Vidal et al. 2001; Martin-Fernandez et al. 2003 ), I propose the following formulation.”
An R package for statistical provenance analysis (2016, background): “Various ways have been proposed to deal with this problem (e.g., Martin-Fernández et al., 2003), but none of these are implemented in provenance (yet).”
Inferring Correlation Networks from Genomic Survey Data (2012, methodology): “We emphasize that the treatment of zero values is a challenge that is in no way unique to compositional data, but is merely highlighted by the log-ratio transformations employed to analyze these data [21].”
Bayesian-multiplicative treatment of count zeros in compositional data sets (2015, methodology): “In addition, according to Martín-Fernández et al. (2003), the non-zero parts in BM (3.1) are modified in a multiplicative way.”