Pour tirer pleinement profit des avantages liés au déploiement dans les nuages, les applications doivent s’exécuter sur des environnements configurés de façon à répondre précisément aux besoins desdites applications. Nous considérons que la sélection et la configuration d’un nuage peut s’appuyer sur une approche basée sur les Lignes de Produits Logiciels (LPLs). Les LPLs sont définies pour exploiter les points communs par la définition d’éléments réutilisables. Cette thèse propose une approche basée sur les LPLs pour sélectionner et configurer un nuage en fonction des besoins d’une application à déployer. Concrètement, nous introduisons un modèle de variabilité permettant de décrire les points communs des différents nuages. Ce modèle de variabilité est notamment étendu avec des attributs et des multiplicités, qui peuvent être également impliqués dans des contraintes complexes. Ensuite, nous proposons une approche permettant de vérifier la cohérence de modèles de variabilité étendus avec des cardinalités, en particulier lorsque ceux-ci évoluent. En cas d’incohérence, nous fournissons un support permettant d’en expliquer son origine et sa cause. Enfin, nous proposons une plateforme automatisée de sélection et configuration de nuages, permettant la dérivation de fichiers de configuration relatifs aux besoins de l’application à déployer en fonction du nuage choisi. Ce travail de recherche s’est effectué dans le cadre du projet européen PaaSage. Les expérimentations menées montrent les avantages de notre approche basée sur les LPLs et, en particulier, comment son utilisation améliore la fiabilité lors d’un déploiement, tout en proposant une plateforme flexible et extensible. / To benefit from the promise of the cloud computing paradigm, applications must be deployed on well-suited and configured cloud environments fulfilling the application’s requirements. We consider that the selection and configuration of such environments can leverage Software Product Line (SPL) principles. SPLs were defined to take advantage of software commonalities through the definition of reusable artifacts. This thesis thus proposes an approach based on SPLs to select and configure cloud environments regarding the requirements related to the application to deploy. In particular, we introduce a variability model enabling the description of commonalities and variabilities between clouds as feature models. In addition, we extend this variability model with attributes and cardinalities, together with constraints over them. Then, we propose an approach to check the consistency of cardinality-based feature models when evolving those models. Our approach provides a support to detect and explain automatically a cardinality inconsistency. Finally, we propose an automated platform to select and configure cloud environments, which generates configuration scripts regarding the requirements of the application to deploy. This work has been done as part of the European PaaSage project. The experiments we conducted to evaluate our approach show that it is well suited to handle the configuration of cloud environments, being both scalable and practical while improving the reliability of the deployment.

Feature models are used for variability management in software product lines. Instead of developing a single feature model, merging small models can be an effective solution to obtain a unified view. Since each stakeholder views the product family from a different perspective, conflicts may occur during merging. In this research, merging of feature models arising from different viewpoints is considered. A normative procedure is proposed to merge feature models by applying local rules. This procedure can merge feature models with cross-tree relationships between sibling features. Application of the local rules is demonstrated with examples.

Software product line engineering is a common method for designing complex software systems. Feature modeling is the most common approach to specify product lines. A feature model is a feature diagram (a special tree of features) plus some crosscutting constraints. Feature modeling languages are grouped into basic and cardinality-based models. The common understanding of the semantics of feature models is a Boolean semantics. We discuss a major deficiency of this semantics and fix it by applying, in turn, modal logic, the theory of multisets, and formal language theory. In order to adequately represent the semantics of basic models, we propose a Kripke semantics and show that basic feature modeling needs a modal rather than Boolean logic. We propose two multiset based theories for cardinality-based feature diagrams, called flat and hierarchical semantics. We show that the hierarchical semantics of a given cardinality-based diagram captures all information in the diagram. We also charac- terize sets of multisets, which can provide a hierarchical semantics of some diagrams. We provide three different reduction processes going from a cardinality-based diagram to an appropriate regular expression. As for crosscutting constraints, we propose a formal language interpretation of them. We also characterize some existing analysis operations over feature models in terms of operations on the corresponding languages and discuss the relevant decidability problems. / Thesis / Doctor of Philosophy (PhD)

In this dissertation we lay the groundwork of automated analysis of extended feature models with constraint programming. Among different proposals, feature modeling has proven to be very effective for modeling and managing variability in Software Product Lines. However, industrial experiences showed that feature models often grow too large with hundreds of features and complex cross-tree relationships, which necessitates automated analysis support. To address this issue we present a mapping from extended feature models, which may include complex feature-feature, feature-attribute and attribute-attribute cross-tree relationships as well as global constraints, to constraint logic programming over finite domains. Then, we discuss the effects of including complex feature attribute relationships on various analysis operations defined on the feature models. As new types of variability emerge due to the inclusion of feature attributes in cross-tree relationships, we discuss the necessity of reformulation of some of the analysis operations and suggest a revised understanding for some other. We also propose new analysis operations arising due to the nature of the new variability introduced. Then we propose a transformation from extended feature models to basic/cardinality-based feature models that may be applied under certain circumstances and enables using SAT or BDD solvers in automated analysis of extended feature models. Finally, we discuss the role of the context information in feature modeling, and propose to use context information in staged configuration of feature-models.

Os meandros da Corrente do Brasil ao largo do sudeste brasileiro s~ao singulares em termos de sua composic¸ ~ao termohalina. Estes v´ ortices, principalmente na regi~ao de Cabo Frio e Cabo de S~ao Tom´e, constituem um sistema integrado a ressurg^encia costeira e a pr´opria Corrente do Brasil, caracterizando um sistema din^amico que propomos denominar Corrente do Brasil-v ´ ortice-ressurg^encia. Apresentamos uma metodologia de construc¸ ~ao de modelos param´etricos das feic¸ ~oes oceanogr´aficas (ou Modelos de Feic¸ ~ao) que comp~oem o sistema. Estes modelos s~ao capazes de construir tridimensionalmente feic¸ ~oes oceanogr´aficas individuais a partir de dados sin´oticos de superf´?cie e fundo para elaborac¸ ~ao de campos iniciais de modelos num´ericos, com o objetivo de compreender a interac¸ ~ao do sistema din^amico Corrente do Brasil-v´ ortice-ressurg^encia. Atrav´es da simulac¸ ~ao num´erica dos campos iniciais baseados nos Modelos de Feic¸ ~ao investigamos o processo que conduz `a assimetria termohalina dos v´ ortices da Corrente do Brasil, acima descrita. Esta foi devida a aproximac¸ ~ao desses v´ ortices junto ao talude continental e ao ajustamento barocl´?nico das correntes `as menores profundidades. Verificamos, ainda, que a penetrac¸a~o da A´ gua Central do Atla^ntico Sul pela camada de Ekman de fundo n~ao tem efeito significativo na estrutura dos v´ ortices. Realizamos experimentosnum´ericos adicionais com o objetivo de investigar quais os tipos de instabilidade geof´?sica estariam envolvidas no fen^omeno do crescimento dos meandros e v´ ortices da Corrente do Brasil. Analisando os resultados atrav´es de c´alculos de convers~ao de energia pudemos concluir que o sistema Corrente do Brasil-v´ ortice-ressurg^encia apresenta instabilidade mista ( barotr ´opica e barocl´?nica). Contudo, verificamos quantitativamente que a instabilidade barocl´?nica foi dominante. O sistema, no entanto, mostrou-se sens´?vel `a ac¸ ~ao da tens~ao de cisalhamento do vento de nordeste na regi~ao e consequentemente `a ressurg^encia costeira. Interpretamos tal sensibilidade como fator respons´avel pela relev^ancia da instabilidade barotr ´opica no processo de crescimento dos meandros da Corrente do Brasil. Por fim, verificamos que um meandro cicl ^onico frontal, quase-estacion´ario, em condic¸ ~oes especiais de crescimento por mecanismo de instabilidade, pode auxiliar no estabelecimento da ressurg^encia costeira. O v´ ortice do Cabo de S~ao Tom´e simulado apresenta velocidade de fase muito baixa e cresce em direc¸ ~ao a oceano aberto, de forma praticamente perpendicular `a quebra de plataforma. Logo, o crescimento do vo´ rtice de Cabo de Sa~o Tome´ faz com que o mesmo advecte A´ gua Costeira para seu interior provocando, por continuidade, afloramento da A´ gua central do Atla^ntico Sul junto `a costa. Assim, podemos dizer que h´a, de fato, uma interac¸ ~ao entre o sistema costeiro e a atividade de meso-escala da Corrente do Brasil. / The Brazil Current meanders off the Brazilian southeast coast are unique in terms of their thermohaline structure. These eddies, especially off the Cape Frio and S~ao Tom´e region constitute an integrated system with the Brazil Current and the coastal upwelling, characterizing a dynamic system which we propose to nominate as Brazil Current-eddy-upwelling system. We present a methodology of parametric model construction of the oceanographic features (or Feature Models) which compose the inferred system . These models are capable of three-dimensionally building independent the relevant features from synoptic surface and bottom data in order to compose the initial fields for numerical models with the objective of understanding the dynamics of the interaction of Brazil Current-eddy-upwelling system. By means of numerical simulation usingthe initial fields based on the Feature Models, we investigated the process which conducts to the thermohaline asymmetry of the eddies in the Brazil Current as mentioned above. This is due to the baroclinic approximation of the currents to shallower depths. The penetration of the Central Water in the South Atlantic by the bottom Ekman layer does not have a significant effect on the eddy structure. Additional numerical experiments were carried out with the objective of unders tanding which of the geophysical instabilities were involved in the growth of the Brazil Current meanders and eddies. The results analyzed through calculation of energy conversion allowed us to conclude that the Brazil Current-eddie-upwelling system presents mixed instability (barotropic and baroclinic). However, the baroclinic instability is dominant and is mostly associated to the vertical shear of the oceanic currents. The system, however,was sensitive to the action of the northeast wind shear and consequently to the coastal upwelling. Such sensibility was interpreted as the main responsible for the relevance of barotropic instability in the growth process of meanders in the Brazil Current. Lastly, we verified that a quasi-stationary frontal cyclonic meander, in special growth conditions by instability mechanisms, mayhelp in establishing coastal upwelling. The simulated S~ao Tom´e Cape Eddy presented a very low phase velocity and grew towards the open sea, almost perpendicularly to the continental shelf break. Therefore, the growth of the S~ao Tom´e Cape Eddy advects Coastal Water into its center and, therefore, lead to upwelling of South Atlantic Central Water near the coast. We can therefore say that there are in fact interaction mechanisms between the coastal system and the meso-scale activity of the Brazil Current.

Os meandros da Corrente do Brasil ao largo do sudeste brasileiro s~ao singulares em termos de sua composic¸ ~ao termohalina. Estes v´ ortices, principalmente na regi~ao de Cabo Frio e Cabo de S~ao Tom´e, constituem um sistema integrado a ressurg^encia costeira e a pr´opria Corrente do Brasil, caracterizando um sistema din^amico que propomos denominar Corrente do Brasil-v ´ ortice-ressurg^encia. Apresentamos uma metodologia de construc¸ ~ao de modelos param´etricos das feic¸ ~oes oceanogr´aficas (ou Modelos de Feic¸ ~ao) que comp~oem o sistema. Estes modelos s~ao capazes de construir tridimensionalmente feic¸ ~oes oceanogr´aficas individuais a partir de dados sin´oticos de superf´?cie e fundo para elaborac¸ ~ao de campos iniciais de modelos num´ericos, com o objetivo de compreender a interac¸ ~ao do sistema din^amico Corrente do Brasil-v´ ortice-ressurg^encia. Atrav´es da simulac¸ ~ao num´erica dos campos iniciais baseados nos Modelos de Feic¸ ~ao investigamos o processo que conduz `a assimetria termohalina dos v´ ortices da Corrente do Brasil, acima descrita. Esta foi devida a aproximac¸ ~ao desses v´ ortices junto ao talude continental e ao ajustamento barocl´?nico das correntes `as menores profundidades. Verificamos, ainda, que a penetrac¸a~o da A´ gua Central do Atla^ntico Sul pela camada de Ekman de fundo n~ao tem efeito significativo na estrutura dos v´ ortices. Realizamos experimentosnum´ericos adicionais com o objetivo de investigar quais os tipos de instabilidade geof´?sica estariam envolvidas no fen^omeno do crescimento dos meandros e v´ ortices da Corrente do Brasil. Analisando os resultados atrav´es de c´alculos de convers~ao de energia pudemos concluir que o sistema Corrente do Brasil-v´ ortice-ressurg^encia apresenta instabilidade mista ( barotr ´opica e barocl´?nica). Contudo, verificamos quantitativamente que a instabilidade barocl´?nica foi dominante. O sistema, no entanto, mostrou-se sens´?vel `a ac¸ ~ao da tens~ao de cisalhamento do vento de nordeste na regi~ao e consequentemente `a ressurg^encia costeira. Interpretamos tal sensibilidade como fator respons´avel pela relev^ancia da instabilidade barotr ´opica no processo de crescimento dos meandros da Corrente do Brasil. Por fim, verificamos que um meandro cicl ^onico frontal, quase-estacion´ario, em condic¸ ~oes especiais de crescimento por mecanismo de instabilidade, pode auxiliar no estabelecimento da ressurg^encia costeira. O v´ ortice do Cabo de S~ao Tom´e simulado apresenta velocidade de fase muito baixa e cresce em direc¸ ~ao a oceano aberto, de forma praticamente perpendicular `a quebra de plataforma. Logo, o crescimento do vo´ rtice de Cabo de Sa~o Tome´ faz com que o mesmo advecte A´ gua Costeira para seu interior provocando, por continuidade, afloramento da A´ gua central do Atla^ntico Sul junto `a costa. Assim, podemos dizer que h´a, de fato, uma interac¸ ~ao entre o sistema costeiro e a atividade de meso-escala da Corrente do Brasil. / The Brazil Current meanders off the Brazilian southeast coast are unique in terms of their thermohaline structure. These eddies, especially off the Cape Frio and S~ao Tom´e region constitute an integrated system with the Brazil Current and the coastal upwelling, characterizing a dynamic system which we propose to nominate as Brazil Current-eddy-upwelling system. We present a methodology of parametric model construction of the oceanographic features (or Feature Models) which compose the inferred system . These models are capable of three-dimensionally building independent the relevant features from synoptic surface and bottom data in order to compose the initial fields for numerical models with the objective of understanding the dynamics of the interaction of Brazil Current-eddy-upwelling system. By means of numerical simulation usingthe initial fields based on the Feature Models, we investigated the process which conducts to the thermohaline asymmetry of the eddies in the Brazil Current as mentioned above. This is due to the baroclinic approximation of the currents to shallower depths. The penetration of the Central Water in the South Atlantic by the bottom Ekman layer does not have a significant effect on the eddy structure. Additional numerical experiments were carried out with the objective of unders tanding which of the geophysical instabilities were involved in the growth of the Brazil Current meanders and eddies. The results analyzed through calculation of energy conversion allowed us to conclude that the Brazil Current-eddie-upwelling system presents mixed instability (barotropic and baroclinic). However, the baroclinic instability is dominant and is mostly associated to the vertical shear of the oceanic currents. The system, however,was sensitive to the action of the northeast wind shear and consequently to the coastal upwelling. Such sensibility was interpreted as the main responsible for the relevance of barotropic instability in the growth process of meanders in the Brazil Current. Lastly, we verified that a quasi-stationary frontal cyclonic meander, in special growth conditions by instability mechanisms, mayhelp in establishing coastal upwelling. The simulated S~ao Tom´e Cape Eddy presented a very low phase velocity and grew towards the open sea, almost perpendicularly to the continental shelf break. Therefore, the growth of the S~ao Tom´e Cape Eddy advects Coastal Water into its center and, therefore, lead to upwelling of South Atlantic Central Water near the coast. We can therefore say that there are in fact interaction mechanisms between the coastal system and the meso-scale activity of the Brazil Current.

Intratumoural heterogeneity complicates the molecular interpretation of biopsies, as multiple distinct tumour genomes are sampled and analysed at once. Ignoring the presence of these populations can lead to erroneous conclusions, and so a correct analysis must account for the clonal structure of the sample. Several methods to reconstruct tumour clonality from sequencing data have been proposed, spanning methods that either do not consider phylogenetic constraints or posit a perfect phylogeny. Models of the first type are typically latent feature models that can describe the observed data flexibly, but whose results may not be reconcilable with a phylogeny. The second type, instead, generally comprises non-parametric mixture models, with strict assumptions on the tumour’s evolutionary process. The focus of this dissertation is on the development of a phylogenetic latent feature model that can bridge the advantages of these two approaches, allowing deviations from a perfect phylogeny. The work is recounted by three statistical models of increasing complexity. First, I present a non-parametric model based on the Indian Buffet Process prior, and highlight the need for phylogenetic constraints. Second, I develop a finite, phylogenetic extension of the previous model, and show that it can outperform competing methods. Third, I generalise the phylogenetic model to arbitrary copy-number states. Markov chain Monte Carlo algorithms are presented to perform inference. The models are tested on datasets that include synthetic data, controlled biological data, and clinical data. In particular, the copy-number generalisation is applied to longitudinal circulating tumour DNA samples. Liquid biopsies that leverage circulating tumour DNA require sensitive techniques in order to detect mutations at low allele fractions. One method that allows sensitive mutation calling is the amplicon sequencing strategy TAm-Seq. I present bioinformatic tools to improve both the development of TAm-Seq amplicon panels and the analysis of its sequencing data. Finally, an enhancement of this method is presented and shown to detect mutations de novo and in a multiplexed manner at allele fractions less than 0.1%.

<p>Software systems have grown larger and more complex in recent years. Generative software development strives to automate software development from a systems family by generating implementations using domain-specific languages. In current practice, specifying domain-specific languages is a manual task requiring expert analysis of multiple information sources. Furthermore, the concepts and relations represented in a language are grown through its usage. Keeping the language consistent with its usage is a time-consuming process requiring manual comparison between the language instances and its language specification. Feature model mining addresses these issues by synthesizing a representative model bottom-up from a sample set of instances called configurations.</p> <p>This thesis presents a mining algorithm that reverse-engineers a probabilistic feature model from a set of individual configurations. A configuration consists of a list of features that are defined as system properties that a stakeholder is interested in. Probabilistic expressions are retrieved from the sample configurations through the use of conjunctive and disjunctive association rule mining. These expressions are used to construct a probabilistic feature model. </p> <p>The mined feature model consists of a hierarchy of features, a set of additional hard constraints and soft constraints. The hierarchy describes the dependencies and alternative relations exhibited among the features. The additional hard constraints are a set of propositional formulas which must be satisfied in a legal configuration. Soft constraints describe likely defaults or common patterns.</p> <p>Systems families are often realized using object-oriented frameworks that provide reusable designs for constructing a family of applications. The mining algorithm is evaluated on a set of applications to retrieve a metamodel of the Java Applet framework. The feature model is then applied to the development of framework-specific modeling languages (FSMLs). FSMLs are domain-specific languages that model the framework-provided concepts and their rules for development.</p> <p>The work presented in this thesis provides the foundation for further research in feature model mining. The strengths and weaknesses of the algorithm are analyzed and the thesis concludes with a discussion of possible extensions.</p>

In Software Product Lines (SPLs), a feature model can be used to represent the similarities and differences within a family of software systems. This allows describing the systems derived from the product line as a unique combination of the features in the model. What makes feature models particularly appealing is the fact that the constraints in the model prevent incompatible features from being part of the same product. Despite the benefits of feature models, constructing and maintaining these models can be a laborious task especially in product lines with a large number of features and constraints. As a result, the study of automated techniques to reason on feature models has become an important research topic in the SPL community in recent years. Two techniques, in particular, have significant appeal for researchers: SAT solvers and Binary Decision Diagrams (BDDs). Each technique has been applied successfully for over four decades now to tackle many practical combinatorial problems in various domains. Currently, several approaches have proposed the compilation of feature models to specific logic representations to enable the use of SAT solvers and BDDs. In this thesis, we argue that several critical issues related to the use of SAT solvers and BDDs have been consistently neglected. For instance, satisfiability is a well-known NP-complete problem which means that, in theory, a SAT solver might be unable to check the satisfiability of a feature model in a feasible amount of time. Similarly, it is widely known that the size of BDDs can become intractable for large models. At the same time, we currently do not know precisely whether these are real issues when feature models, especially large ones, are compiled to SAT and BDD representations. Therefore, in our research we provide a significant step forward in the state-of-the-art by examining deeply many relevant properties of the feature modeling domain and the mechanics of SAT solvers and BDDs and the sensitive issues related to these techniques when applied in that domain. Specifically, we provide more accurate explanations for the space and/or time (in)tractability of these techniques in the feature modeling domain, and enhance the algorithmic performance of these techniques for reasoning on feature models. The contributions of our work include the proposal of novel heuristics to reduce the size of BDDs compiled from feature models, several insights on the construction of efficient domain-specific reasoning algorithms for feature models, and empirical studies to evaluate the efficiency of SAT solvers in handling very large feature models.

<p>Software systems have grown larger and more complex in recent years. Generative software development strives to automate software development from a systems family by generating implementations using domain-specific languages. In current practice, specifying domain-specific languages is a manual task requiring expert analysis of multiple information sources. Furthermore, the concepts and relations represented in a language are grown through its usage. Keeping the language consistent with its usage is a time-consuming process requiring manual comparison between the language instances and its language specification. Feature model mining addresses these issues by synthesizing a representative model bottom-up from a sample set of instances called configurations.</p> <p>This thesis presents a mining algorithm that reverse-engineers a probabilistic feature model from a set of individual configurations. A configuration consists of a list of features that are defined as system properties that a stakeholder is interested in. Probabilistic expressions are retrieved from the sample configurations through the use of conjunctive and disjunctive association rule mining. These expressions are used to construct a probabilistic feature model. </p> <p>The mined feature model consists of a hierarchy of features, a set of additional hard constraints and soft constraints. The hierarchy describes the dependencies and alternative relations exhibited among the features. The additional hard constraints are a set of propositional formulas which must be satisfied in a legal configuration. Soft constraints describe likely defaults or common patterns.</p> <p>Systems families are often realized using object-oriented frameworks that provide reusable designs for constructing a family of applications. The mining algorithm is evaluated on a set of applications to retrieve a metamodel of the Java Applet framework. The feature model is then applied to the development of framework-specific modeling languages (FSMLs). FSMLs are domain-specific languages that model the framework-provided concepts and their rules for development.</p> <p>The work presented in this thesis provides the foundation for further research in feature model mining. The strengths and weaknesses of the algorithm are analyzed and the thesis concludes with a discussion of possible extensions.</p>