Nicolas S. Müller’s research while affiliated with University of Geneva and other places

Collection of ancillary functions and utilities to be used in conjunction with the 'TraMineR' package for sequence data exploration. Includes, among others, specific functions such as state survival plots, position-wise group-typical states, dynamic sequence indicators, and dissimilarities between event sequences. Also includes contributions by non-members of the TraMineR team such as the relative frequency plot and methods for polyadic data.

Background: Most of the existing research relating to the life courses of people with psychiatric symptoms focuses on the occurrence and the impact of non-normative events on the onsets of crises; it usually disregards the more regular dimensions of life, such as work, family and intimate partnerships that may be related to the timing and seriousness of psychiatric problems. An additional reason for empirically addressing life trajectories of individuals with psychiatric problems relates to recent changes of family and occupational trajectories in relation to societal trends such as individualization and pluralization of life courses. Aim: This paper explores the life trajectories of 86 individuals under clinical supervision and proposes a typology of their occupational, co-residence and intimacy trajectories. The results are discussed in light of the life-course paradigm. Method: A multidimensional optimal matching analysis was performed on a sample of 86 individuals under clinical supervision to create a typology of trajectories. The influence of these trajectories on psychiatric disorders, evaluated using a SCL-90-R questionnaire, was then assessed using linear regression modelling. Results: The typologies of trajectories showed that the patients developed a diversity of life trajectories. Individuals who have developed a standard life course with few institutionalization periods reported more symptoms and distress than individuals with an institutionalized life trajectory. Conclusion: The results of this study stress that psychiatric patients are social actors who are influenced by society at large and its ongoing process of change. Therefore, it is essential to take into account the diversity of occupational and family trajectories when dealing with individuals in therapeutic settings.

In this article, the authors define a methodological framework for analyzing the relationship between state sequences and covariates. Inspired by the principles of analysis of variance, this approach looks at how the covariates explain the discrepancy of the sequences. The authors use the pairwise dissimilarities between sequences to determine the discrepancy, which makes it possible to develop a series of statistical significance-based analysis tools. They introduce generalized simple and multifactor discrepancy-based methods to test for differences between groups, a pseudo-R2 for measuring the strength of sequence-covariate associations, a generalized Levene statistic for testing differences in the within-group discrepancies, as well as tools and plots for studying the evolution of the differences along the time frame and a regression tree method for discovering the most significant discriminant covariates and their interactions. In addition, the authors extend all methods to account for case weights. The scope of the proposed methodological framework is illustrated using a real-world sequence data set.

This article describes the many capabilities offered by the TraMineR toolbox for categorical sequence data. It focuses more specifically on the analysis and rendering of state sequences. Addressed features include the description of sets of sequences by means of transversal aggregated views, the computation of longitudinal characteristics of individual sequences and the measure of pairwise dissimilarities. Special emphasis is put on the multiple ways of visualizing sequences. The core element of the package is the state se- quence object in which we store the set of sequences together with attributes such as the alphabet, state labels and the color palette. The functions can then easily retrieve this information to ensure presentation homogeneity across all printed and graphical displays. The article also demonstrates how TraMineR’s outcomes give access to advanced analyses such as clustering and statistical modeling of sequence data.

This is a User's Guide of the R package TraMineR version 1.8.

This paper is concerned with the summarization of a set of categorical sequences. More specifically, the problem studied is the determination of the smallest possible number of representative sequences that ensure a given coverage of the whole set, i.e. that have together a given percentage of sequences in their neighbourhood. The proposed heuristic for extracting the representative subset requires as main arguments a pairwise distance matrix, a representativeness criterion and a distance threshold under which two sequences are considered as redundant or, identically, in the neighborhood of each other. It first builds a list of candidates using a representativeness score and then eliminates redundancy. We propose also a visualization tool for rendering the results and quality measures for evaluating them. The proposed tools have been implemented in our TraMineR R package for mining and visualizing sequence data and we demonstrate their efficiency on a real world example from social sciences. The methods are nonetheless by no way limited to social science data and should prove useful in many other domains. KeywordsCategorical sequences–Representatives–Pairwise dissimilarities–Discrepancy of sequences–Summarizing sets of sequences–Visualization

La recherche de liens entre objets fréquents a été popularisée par les méthodes d’extraction de règles d’association. Dans le cas de séquences d’événements, les méthodes de fouille permettent d’extraire des sous-séquences qui peuvent ensuite être exprimées sous la forme de règles d’association séquentielle entre événements. Cette utilisation de la fouille de séquences pour la recherche de liens entre des événements pose deux problèmes. Premièrement, le critère principal utilisé pour sélectionner les sous-séquences d’événements est la fréquence, or les occurrences de certains événements peuvent être fortement liées entre elles même lorsqu’elles sont peu fréquentes. Deuxièmement, les mesures actuelles utilisées pour caractériser les règles d’association ne tiennent pas compte du caractère temporel des données, comme l’importance du timing des événements ou le problème des données censurées. Dans cet article, nous proposons une méthode pour rechercher des liens significatifs entre des événements à l’aide de modèles de durée. Les règles d’association sont construites à partir des motifs séquentiels observés dans un ensemble de séquences. L’influence sur le risque que l’événement « conclusion » se produise après le ou les événements « prémisse » est estimée à l’aide d’un modèle semi-paramétrique à risques proportionnels. Outre la présentation de la méthode, l’article propose une comparaison avec d’autres mesures d’association.