Generalization, Combination and Extension of Functional Clustering Algorithms: The <i>R</i> Package <b>funcy</b>

Yassouridis, Christina; Ernst, Dominik; Leisch, Friedrich

doi:10.18637/jss.v085.i09

Cited by 5 publications

(5 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Clustering is then performed on the basis coefficients rather than on multivariate observations, which has the benefit of temporal structures being conserved. Analyses are performed using the method fscm from the R-package funcy (Yassouridis et al, 2016), which applies the functional mixed mixture model of Jiang and Serban (2012) to perform the clustering. The 2003 and 2015 events are analyzed separately, based on monthly mean discharges of the January-October period.…”

Section: Functional Clusteringmentioning

confidence: 99%

The European 2015 drought from a hydrological perspective

Laaha

Gauster

Tallaksen

et al. 2017

Hydrol. Earth Syst. Sci.

172

152

View full text Add to dashboard Cite

Abstract. In 2015 large parts of Europe were affected by drought. In this paper, we analyze the hydrological footprint (dynamic development over space and time) of the drought of 2015 in terms of both severity (magnitude) and spatial extent and compare it to the extreme drought of 2003. Analyses are based on a range of low flow and hydrological drought indices derived for about 800 streamflow records across Europe, collected in a community effort based on a common protocol. We compare the hydrological footprints of both events with the meteorological footprints, in order to learn from similarities and differences of both perspectives and to draw conclusions for drought management. The region affected by hydrological drought in 2015 differed somewhat from the drought of 2003, with its center located more towards eastern Europe. In terms of low flow magnitude, a region surrounding the Czech Republic was the most affected, with summer low flows that exhibited return intervals of 100 years and more. In terms of deficit volumes, the geographical center of the event was in southern Germany, where the drought lasted a particularly long time. A detailed spatial and temporal assessment of the 2015 event showed that the particular behavior in these regions was partly a result of diverging wetness preconditions in the studied catchments. Extreme droughts emerged where preconditions were particularly dry. In regions with wet preconditions, low flow events developed later and tended to be less severe. For both the 2003 and 2015 events, the onset of the hydrological drought was well correlated with the lowest flow recorded during the event (low flow magnitude), pointing towards a potential for early warning of the severity of streamflow drought. Time series of monthly drought indices (both streamflow- and climate-based indices) showed that meteorological and hydrological events developed differently in space and time, both in terms of extent and severity (magnitude). These results emphasize that drought is a hazard which leaves different footprints on the various components of the water cycle at different spatial and temporal scales. The difference in the dynamic development of meteorological and hydrological drought also implies that impacts on various water-use sectors and river ecology cannot be informed by climate indices alone. Thus, an assessment of drought impacts on water resources requires hydrological data in addition to drought indices based solely on climate data. The transboundary scale of the event also suggests that additional efforts need to be undertaken to make timely pan-European hydrological assessments more operational in the future.

show abstract

Section: Functional Clusteringmentioning

confidence: 99%

The European 2015 drought from a hydrological perspective

Laaha

Gauster

Tallaksen

et al. 2017

Hydrol. Earth Syst. Sci.

172

152

View full text Add to dashboard Cite

show abstract

“…In this work functional clustering (FC) was performed by the k ‐centres method described by Chiou and Li (2007) and implemented in the funcit function (Funcy R package v. 0.8.6; Yassouridis et al, 2018). This method is a functional counterpart of the k ‐means algorithm as cluster membership is predicted with a reclassification step: alternately, curves are assigned to classes, and classes are calculated anew depending on their assigned curves.…”

Section: Methodsmentioning

confidence: 99%

“…In this work functional clustering (FC) was performed by the kcentres method described by Chiou and Li (2007) and implemented in the funcit function (Funcy R package v. 0.8.6; Yassouridis et al, 2018).…”

Section: Characterization -Functional Clusteringmentioning

confidence: 99%

A statistical protocol to describe differences among nutrient utilization patterns of Fusarium spp. and Trichoderma gamsii

et al. 2021

View full text Add to dashboard Cite

The Biolog phenotype microarrays (PM) system offers a simple and cheap tool to rapidly provide a high throughput of information about the phenotypes of fungal isolates in a short time. In order to improve the use of the PM system in fungal ecology studies, the present work proposes a new statistical protocol based on two approaches, that is, a functional principal components analysis to describe similarity patterns of growth curves, and a Bayesian generalized additive model (GAM) to allow inferences on specific growth features, in order to analyse nutrient fungal utilization in a model system including four causal agents of Fusarium head blight, the natural competitor Fusarium oxysporum, and the beneficial isolate Trichoderma gamsii T6085. Analysis of data collected by the Biolog PM in our biological system showed a different nutritional competitive potential of the four pathogens, as well as an intermediate behaviour of the natural competitor and of our biocontrol agent. This protocol, applicable to different fungal phenotypical studies at both isolate and community level, allows a full exploitation of data obtained by the PM system and provides important information about the nutritional pattern of a single isolate compared to those of other fungi, a key factor to be exploited in biocontrol strategies.

show abstract

“…The empirical analysis in this article is conducted using R 4.2.1. In each experiment, the distclust, iterSubspace [20], funclust [21], funHDDC [22], fscm [23], and waveclust [24] algorithms from the Funcy library (version 1.0.0) [25] are compared. The functional data clustering algorithms based on the Mahalanobis distance are designed by improving the K-means clustering algorithm.…”

Section: Data Sources and Experimental Settingsmentioning

confidence: 99%

Functional Data Clustering Via Functional Mahalanobis Distance

Zao

2023

HSET

View full text Add to dashboard Cite

As an exploratory data analysis method, functional data clustering aims to identify the underlying features of the observed data. In this context, this paper proposes a functional data clustering method based on functional Mahalanobis distance. As a distance-based non-parametric clustering model, the proposed method can effectively avoid the disadvantages of generative models and has excellent properties of decoupling and dimension standardization. Compared with other functional data clustering models, this method has lower computational complexity. In addition, the proposed method can be applied to any distance-based multivariate clustering method, thus generalizing it to the case of functional data. In practical data analysis, this paper compares the performance of this method with some other functional clustering methods, using k-means clustering as an example, and finds that it has better performance in terms of purity, adjusted Rand index, and computational speed. Finally, the idea of using Mahalanobis distance for functional data distance measurement can also be extended to construct kernel functions for measuring similarity between functional data samples, thus developing non-linear functional data analysis methods based on reproducing kernel theory.

show abstract

Generalization, Combination and Extension of Functional Clustering Algorithms: The R Package funcy

Cited by 5 publications

References 16 publications

The European 2015 drought from a hydrological perspective

The European 2015 drought from a hydrological perspective

A statistical protocol to describe differences among nutrient utilization patterns of Fusarium spp. and Trichoderma gamsii

Functional Data Clustering Via Functional Mahalanobis Distance

Contact Info

Product

Resources

About