Automatic evaluation and data generation for analytical chemistry instrumental analysis exercises

In general, laboratory activities are costly in terms of time, space, and money. As such, the ability to provide realistically simulated laboratory data that enables students to practice data analysis techniques as a complementary activity would be expected to reduce these costs while opening up very interesting possibilities. In the present work, a novel methodology is presented for design of analytical chemistry instrumental analysis exercises that can be automatically personalized for each student and the results evaluated immediately. The proposed system provides each student with a different set of experimental data generated randomly while satisfying a set of constraints, rather than using data obtained from actual laboratory work. This allows the instructor to provide students with a set of practical problems to complement their regular laboratory work along with the corresponding feedback provided by the system's automatic evaluation process. To this end, the Goodle Grading Management System (GMS), an innovative web-based educational tool for automating the collection and assessment of practical exercises for engineering and scientific courses, was developed. The proposed methodology takes full advantage of the Goodle GMS fusion code architecture. The design of a particular exercise is provided ad hoc by the instructor and requires basic Matlab knowledge. The system has been employed with satisfactory results in several university courses. To demonstrate the automatic evaluation process, three exercises are presented in detail. The first exercise involves a linear regression analysis of data and the calculation of the quality parameters of an instrumental analysis method. The second and third exercises address two different comparison tests, a comparison test of the mean and a t-paired test.

Generación de Casos de Prueba en Programación Orientada a Objetos = Test Case Generation in Object-Oriented Programming

Las pruebas de software (Testing) son en la actualidad la técnica más utilizada para la validación y la evaluación de la calidad de un programa. El testing está integrado en todas las metodologías prácticas de desarrollo de software y juega un papel crucial en el éxito de cualquier proyecto de software. Desde las unidades de código más pequeñas a los componentes más complejos, su integración en un sistema de software y su despliegue a producción, todas las piezas de un producto de software deben ser probadas a fondo antes de que el producto de software pueda ser liberado a un entorno de producción. La mayor limitación del testing de software es que continúa siendo un conjunto de tareas manuales, representando una buena parte del coste total de desarrollo. En este escenario, la automatización resulta fundamental para aliviar estos altos costes. La generación automática de casos de pruebas (TCG, del inglés test case generation) es el proceso de generar automáticamente casos de prueba que logren un alto recubrimiento del programa. Entre la gran variedad de enfoques hacia la TCG, esta tesis se centra en un enfoque estructural de caja blanca, y más concretamente en una de las técnicas más utilizadas actualmente, la ejecución simbólica. En ejecución simbólica, el programa bajo pruebas es ejecutado con expresiones simbólicas como argumentos de entrada en lugar de valores concretos. Esta tesis se basa en un marco general para la generación automática de casos de prueba dirigido a programas imperativos orientados a objetos (Java, por ejemplo) y basado en programación lógica con restricciones (CLP, del inglés constraint logic programming). En este marco general, el programa imperativo bajo pruebas es primeramente traducido a un programa CLP equivalente, y luego dicho programa CLP es ejecutado simbólicamente utilizando los mecanismos de evaluación estándar de CLP, extendidos con operaciones especiales para el tratamiento de estructuras de datos dinámicas. Mejorar la escalabilidad y la eficiencia de la ejecución simbólica constituye un reto muy importante. Es bien sabido que la ejecución simbólica resulta impracticable debido al gran número de caminos de ejecución que deben ser explorados y a tamaño de las restricciones que se deben manipular. Además, la generación de casos de prueba mediante ejecución simbólica tiende a producir un número innecesariamente grande de casos de prueba cuando es aplicada a programas de tamaño medio o grande. Las contribuciones de esta tesis pueden ser resumidas como sigue. (1) Se desarrolla un enfoque composicional basado en CLP para la generación de casos de prueba, el cual busca aliviar el problema de la explosión de caminos interprocedimiento analizando de forma separada cada componente (p.ej. método) del programa bajo pruebas, almacenando los resultados y reutilizándolos incrementalmente hasta obtener resultados para el programa completo. También se ha desarrollado un enfoque composicional basado en especialización de programas (evaluación parcial) para la herramienta de ejecución simbólica Symbolic PathFinder (SPF). (2) Se propone una metodología para usar información del consumo de recursos del programa bajo pruebas para guiar la ejecución simbólica hacia aquellas partes del programa que satisfacen una determinada política de recursos, evitando la exploración de aquellas partes del programa que violan dicha política. (3) Se propone una metodología genérica para guiar la ejecución simbólica hacia las partes más interesantes del programa, la cual utiliza abstracciones como generadores de trazas para guiar la ejecución de acuerdo a criterios de selección estructurales. (4) Se propone un nuevo resolutor de restricciones, el cual maneja eficientemente restricciones sobre el uso de la memoria dinámica global (heap) durante ejecución simbólica, el cual mejora considerablemente el rendimiento de la técnica estándar utilizada para este propósito, la \lazy initialization". (5) Todas las técnicas propuestas han sido implementadas en el sistema PET (el enfoque composicional ha sido también implementado en la herramienta SPF). Mediante evaluación experimental se ha confirmado que todas ellas mejoran considerablemente la escalabilidad y eficiencia de la ejecución simbólica y la generación de casos de prueba. ABSTRACT Testing is nowadays the most used technique to validate software and assess its quality. It is integrated into all practical software development methodologies and plays a crucial role towards the success of any software project. From the smallest units of code to the most complex components and their integration into a software system and later deployment; all pieces of a software product must be tested thoroughly before a software product can be released. The main limitation of software testing is that it remains a mostly manual task, representing a large fraction of the total development cost. In this scenario, test automation is paramount to alleviate such high costs. Test case generation (TCG) is the process of automatically generating test inputs that achieve high coverage of the system under test. Among a wide variety of approaches to TCG, this thesis focuses on structural (white-box) TCG, where one of the most successful enabling techniques is symbolic execution. In symbolic execution, the program under test is executed with its input arguments being symbolic expressions rather than concrete values. This thesis relies on a previously developed constraint-based TCG framework for imperative object-oriented programs (e.g., Java), in which the imperative program under test is first translated into an equivalent constraint logic program, and then such translated program is symbolically executed by relying on standard evaluation mechanisms of Constraint Logic Programming (CLP), extended with special treatment for dynamically allocated data structures. Improving the scalability and efficiency of symbolic execution constitutes a major challenge. It is well known that symbolic execution quickly becomes impractical due to the large number of paths that must be explored and the size of the constraints that must be handled. Moreover, symbolic execution-based TCG tends to produce an unnecessarily large number of test cases when applied to medium or large programs. The contributions of this dissertation can be summarized as follows. (1) A compositional approach to CLP-based TCG is developed which overcomes the inter-procedural path explosion by separately analyzing each component (method) in a program under test, stowing the results as method summaries and incrementally reusing them to obtain whole-program results. A similar compositional strategy that relies on program specialization is also developed for the state-of-the-art symbolic execution tool Symbolic PathFinder (SPF). (2) Resource-driven TCG is proposed as a methodology to use resource consumption information to drive symbolic execution towards those parts of the program under test that comply with a user-provided resource policy, avoiding the exploration of those parts of the program that violate such policy. (3) A generic methodology to guide symbolic execution towards the most interesting parts of a program is proposed, which uses abstractions as oracles to steer symbolic execution through those parts of the program under test that interest the programmer/tester most. (4) A new heap-constraint solver is proposed, which efficiently handles heap-related constraints and aliasing of references during symbolic execution and greatly outperforms the state-of-the-art standard technique known as lazy initialization. (5) All techniques above have been implemented in the PET system (and some of them in the SPF tool). Experimental evaluation has confirmed that they considerably help towards a more scalable and efficient symbolic execution and TCG.

Test vector generation for the phase II ATLAS event filter trigger upgrade

In the near future, the LHC experiments will continue to be upgraded as the LHC luminosity will increase from the design 1034 to 7.5 × 1034, with the HL-LHC project, to reach 3000 × f b−1 of accumulated statistics. After the end of a period of data collection, CERN will face a long shutdown to improve overall performance by upgrading the experiments and implementing more advanced technologies and infrastructures. In particular, ATLAS will upgrade parts of the detector, the trigger, and the data acquisition system. It will also implement new strategies and algorithms for processing and transferring the data to the final storage. This PhD thesis presents a study of a new pattern recognition algorithm to be used in the trigger system, which is a software designed to provide the information necessary to select physical events from background data. The idea is to use the well-known Hough Transform mathematical formula as an algorithm for detecting particle trajectories. The effectiveness of the algorithm has already been validated in the past, independently of particle physics applications, to detect generic shapes in images. Here, a software emulation tool is proposed for the hardware implementation of the Hough Transform, to reconstruct the tracks in the ATLAS Trigger and Data Acquisition system. Until now, it has never been implemented on electronics in particle physics experiments, and as a hardware implementation it would provide overall latency benefits. A comparison between the simulated data and the physical system was performed on a Xilinx UltraScale+ FPGA device.

Synthetic data generation for the assessment of antimicrobial resistance through machine learning

As a consequence of the diffusion of next generation sequencing techniques, metagenomics databases have become one of the most promising repositories of information about features and behavior of microorganisms. One of the subjects that can be studied from those data are bacteria populations. Next generation sequencing techniques allow to study the bacteria population within an environment by sampling genetic material directly from it, without the needing of culturing a similar population in vitro and observing its behavior. As a drawback, it is quite complex to extract information from those data and usually there is more than one way to do that; AMR is no exception. In this study we will discuss how the quantified AMR, which regards the genotype of the bacteria, can be related to the bacteria phenotype and its actual level of resistance against the specific substance. In order to have a quantitative information about bacteria genotype, we will evaluate the resistome from the read libraries, aligning them against CARD database. With those data, we will test various machine learning algorithms for predicting the bacteria phenotype. The samples that we exploit should resemble those that could be obtained from a natural context, but are actually produced by a read libraries simulation tool. In this way we are able to design the populations with bacteria of known genotype, so that we can relay on a secure ground truth for training and testing our algorithms.

Inferring transport characteristics in a fractured rock aquifer by combining single-hole ground-penetrating radar reflection monitoring and tracer test data

Investigations of solute transport in fractured rock aquifers often rely on tracer test data acquired at a limited number of observation points. Such data do not, by themselves, allow detailed assessments of the spreading of the injected tracer plume. To better understand the transport behavior in a granitic aquifer, we combine tracer test data with single-hole ground-penetrating radar (GPR) reflection monitoring data. Five successful tracer tests were performed under various experimental conditions between two boreholes 6 m apart. For each experiment, saline tracer was injected into a previously identified packed-off transmissive fracture while repeatedly acquiring single-hole GPR reflection profiles together with electrical conductivity logs in the pumping borehole. By analyzing depth-migrated GPR difference images together with tracer breakthrough curves and associated simplified flow and transport modeling, we estimate (1) the number, the connectivity, and the geometry of fractures that contribute to tracer transport, (2) the velocity and the mass of tracer that was carried along each flow path, and (3) the effective transport parameters of the identified flow paths. We find a qualitative agreement when comparing the time evolution of GPR reflectivity strengths at strategic locations in the formation with those arising from simulated transport. The discrepancies are on the same order as those between observed and simulated breakthrough curves at the outflow locations. The rather subtle and repeatable GPR signals provide useful and complementary information to tracer test data acquired at the outflow locations and may help us to characterize transport phenomena in fractured rock aquifers.

A copula-based method for synthetic microarray data generation

In this work, we propose a copula-based method to generate synthetic gene expression data that account for marginal and joint probability distributions features captured from real data. Our method allows us to implant significant genes in the synthetic dataset in a controlled manner, giving the possibility of testing new detection algorithms under more realistic environments.

Synchronous machines parameters identification using load rejection test data

This work shows a computational methodology for the determination of synchronous machines parameters using load rejection test data. The quadrature axis parameters are obtained with a rejection under an arbitrary reference, reducing the present difficulties.

Identification of synchronous machine parameters using load rejection test data

This work shows a computational methodology for the determination of synchronous machines parameters using load rejection test data. By machine modeling one can obtain the quadrature parameters through a load rejection under an arbitrary reference, reducing the present difficulties. The proposed method is applied to a real machine.