Partial evaluation is a program transformation technique that specializes a program with respect to part of its input. While the specialization is typically performed ahead-of-time, moving it to a later stage may expose additional opportunities and allow for faster residual programs to be constructed. In this thesis, we present a method for specializing programs at run-time, for compiled code, using an online partial evaluator. Although partial evaluation has several applications, the evaluation of the method primarily focuses on its performance benefits. The main research problem addressed in this thesis is that of incorporating an online partial evaluator in compiled code. The partial evaluator is a sourceto-source translator that takes and produces an abstract syntax tree (AST). Our approach consists of three parts, namely that of partially evaluating, obtaining a partially evaluable representation and run-time code emitting. Concretely, we use the concept of lifting to store an AST in the compiled code that the partial evaluator then specializes at run-time. The residual code is thereafter naively just-in-time (JIT) compiled through dynamically linking it back to the executable as a shared library. We evaluate the method on several programs and show that the specialized programs sometimes are faster even with a low recursion depth. Though, while the results are promising, the overhead is typically significant and therefore the break-even points are large. Further research, for example using an efficient JIT compiler, is required to better evaluate the performance benefits of the approach. / Partiell evaluering är en programtransformationsteknik som specialiserar ett program givet delar av dess indata. Typisk sätt specialiseras program innan de exekveras, men genom att flytta specialisering till då programmet körs kan ytterligare information utnyttjas och därmed snabbare residualprogram konstrueras. I det här examensarbetet presenteras en metod för att specialisera program i körtid med online partiell evaluering, specifikt för kompilerade program. Metoden utvärderas främst utefter prestanda, men det ska nämnas att partiell evaluering har fler tillämpningar än så. Det huvudsakliga problemet som examensarbetet undersöker är inkorporeringen av en programspecialiserare (partial evaluator) i kompilerad kod. Den programspecialiserare som används tar både som indata och producerar ett abstrakt syntaxträd (AST). Vårt tillvägagångssätt består av tre delar, nämligen programspecialisering, erhållning av en representation som kan specialiseras och slutligen kodgenerering i körtid. Mer specifikt används konceptet lyftning för att spara ett AST i den kompilerade koden som därefter partiellt evalueras av programspecialiseraren under körtid. Som ett sista steg just-in-time (JIT) kompileras residualprogrammet. Detta görs på ett naivt vis genom att programmet kompileras till ett delat bibliotek som därefter dynamiskt länkas tillbaka till huvudprogrammet. Metoden utvärderas på flera program och vi visar att de specialiserade programmen i vissa fall var snabbare och det även med små rekursionsdjup. Resultaten är lovande, men den overhead som metoden ger upphov till är ofta signifikant vilket gör att det krävs många iterationer innan det specialiserade programmet blir snabbare. Ytterligare forskning och tester, till exempel med en effektiv JIT kompilator, är nödvändig för att bättre kunna utvärdera metodens prestandafördelar.

Predicting noise using computational methods is a common approach in the field of acoustics, offering flexibility and results before a potentially costly installation. This project explores the possibility to calculate the change in noise level emitted from a speaker covered by a capsule, filled with porous material and positioned in two different ways, using the finite element method in the program COMSOL for low to mid frequency bands. The Helmholtz equation is solved in the air medium of the model excluding wave propagation in the solid domain of the capsule where it is present. To replicate the attenuation from porous materials in COMSOL the Delany-BazleyMiki poroacoustics model is used. Comparing the calculation to a control measurement conducted at the Marcus Wallenberg Laboratory, provisional predictions of accuracy is given in groups of configurations and in 1/3-octave bands. Using the available computational power in a common office computer it might be difficult to accurately portray the real-world installation since it might be necessary to alter geometric region, physics or frequency rangedue to runtime issues resulting in deviating results. / Att förutspå ljuddämpning med hjälp av beräkningsmetoder är ett vanligt arbetssätt inom akustik som bidrar till flexibilitet och resultat innan en potentiellt kostsam installation. Det här projektet utforskar möjligheten att beräkna förändringen i ljudnivå från en högtalare som är täckt utav en kapsel fylld utav poröst material. Kapseln är placerad i två olika konfigurationer och finita element metoden används genom programmet comsol för att räkna ut ljudnivåer i låg och medelhöga frekvensband. Helmholtz ekvation löses i luftmediet som omger kapseln och högtalaren och Delany-Bazleys modell för porösa material används, vilken finns tillgänglig i COMSOL. Jämförelsen mellan beräkningen och en kontrollmätning gjord i Marcus Wallenbergs Laboratorium ger indikation på precisionen hos modellen, vilket visas i tersband. Med beräkningskraften hos en konventionell kontorsdator kan det vara svårt att noggrant beskriva den verkliga installationen där avsteg från verkligheten ofta handlar om förenklad geometri och fysik eller förändrat frekvensspann för att förhindra eskalerande problem med programmets körtid vilket leder till avvikande resultat.

Yes / A recent pilot study by Chesham et al. in BMC Medicine established some initial effects of the Daily Mile™ using a quasi-experimental repeated measures design, with valid and reliable outcome assessments for moderate-to-vigorous physical activity, fitness and body composition. Their contribution is important and welcome, yet, alone, it is insufficient to justify the recent UK-wide adoption of the Daily Mile within the Childhood Obesity Plan. The study concluded that the Daily Mile had positive effects on moderate-to-vigorous physical activity, fitness and body composition, suggesting that intervention effectiveness was confirmed. However, only some of the significant limitations of the work were addressed. Herein, we identify and discuss six key limitations, which, combined, suggest a more tentative conclusion. In summary, evidence supporting the effectiveness of the Daily Mile is in its infancy and requires refinement to fully justify its widespread adoption. Further, we need to be cautious considering that the full range of its impacts, both positive and negative, remain to be fully established.

Objective: Training and abilities of junior athletes Prepared by: Bc. Petr Dubský Supervisor: PaedDr. Jitka Vindušková, CSc. The aim of the work: The aim of the work was to find out the level of performance preconditions of two junior runners on the middle tracks, analyze the structure and content of their training, their performance in selected tests and the development of performance in their main disciplines. Methodology: The work has the character of a two-case study. Methods of literature search, analysis, comparison and evaluation were used. The analysis of training records and competition results of two athletes from the 2017 season was used to compile this work. I analyzed their performances in individual seasons. I also performed an analysis of training records, where general and special training indicators were evaluated and compared with the results of similar analyzes of other athletes. I analyzed the results of two tests in the biochemical laboratory, which we completed during the preparation for the 2017 season. Conclusion: When comparing the training records with the professional literature and with selected athletes, we come to the conclusion that the training load was insufficient for the monitored athletes. As a result, they could not achieve better results this season. By...

雙次抽樣平均數和變異數管制圖設計之研究 / Control charts are effective tools for detecting manufacturing processes and service processes. Nowadays, much of the data in manufacturing or service industries comes from processes having non-normal or unknown distributions. The commonly used Shewhart control charts, which depend heavily on the normality assumption, are not appropriately used for this situation. In this paper, we propose a standardized dynamic double sampling asymmetric EWMA mean control chart (SDDS EWMA-AM chart), a standardized dynamic double sampling asymmetric EWMA variance control chart (SDDS EWMA-AV chart), and their combined charts (joint SDDS EWMA-AM and SDDS EWMA-AV charts) to monitor process mean, variance and both shifts, respectively. The charts based on the double sampling procedure and two simple distribution-free transformed statistics are used for non-normal distribution of a quality variable. The performance of the proposed charts and that of some existing distribution-free mean and variance charts are compared. Further, a non-normal service times example from the service system of a bank branch is used to illustrate the applications of the proposed charts and to compare detection performance with the existing distribution-free mean and variance control charts. The charts we proposed show superior detection performance compared to the existing distribution-free mean and variance charts. Thus they are recommended.

Thesis (MEng)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: Sedimentation and the effects it has on turbine blades was the primary problem identified in run-of-river (RoR) hydropower schemes. Sedimentation in RoR hydropower schemes also increases trash rack blockage and reduces energy output in the long-term. Damage occurs to all underwater parts that come into contact with sediment. The main concern is sediment passing through the hydropower intake and causing turbine damage. The reason for the abrasion and cavitation of turbine blades is increased sediment loads in river channels. This problem can be overcome in two ways. The first is the use of existing lakes or reservoir storage upstream as natural sand traps, and the second is by investigating the three features associated with river bend diversion, which are: the optimum diversion location in a river bend to minimise the abstraction of sediment, the optimum diversion structure angle to limit coarse sediment diversion, and the sediment load diverted through the intake. The first objective of the research was investigated by construction of a physical model of a curved river channel to determine the location of the deepest scour forming on the outside of the bend. The second objective was to test the diversion orientation to maximize the local scour and thereby limiting sediment diversion at the intake. A third objective was to compare mathematical 2D model simulated scour results with the findings of the laboratory tests to evaluate the reliability of the numerical model predictions. Finally different diverted discharge ratios were tested with different intake setups in the physical model, to evaluate the sediment load diverted. . The first experiment in the curved laboratory channel was to predict where the deepest scour takes place without a diversion structure. This was then followed by placing a diversion structure at the maximum scour position, retrieved from experiment one, and by angling the structure with reference to the flow direction. The flow direction vector was placed as a tangent to the bend and orientated at angles of 0⁰, 30⁰, 45⁰ and 60⁰ into the bend direction. The optimum diversion location was found to be positioned on the outside of the bend, approximately 60⁰ into the channel bend. The final position of maximum scour in a 90⁰ bend corresponds with the Sediment Committee and the Chinese Hydraulic Engineering Societies (1992) prediction of 60⁰ into the bend. The optimal diversion had a 30⁰ angle to the flow direction, as this presented the most efficient and effective scouring in front of the model intake. Numerical simulations were performed with the CCHE 2D (hydrodynamics and sediment dynamics) modelling program. The numerical results were compared to the physical results to validate CCHE as a beneficial simulation tool. It was found that the numerical model predicted the scour depths at the intakes tested with an accuracy of 43.8%, which is within the accuracy range of the sediment transport equation used by the numerical model. The final experiment was the diversion of sediment with different intake level heights and discharges. It was evident from the results that low sediment diversion ratios were achieved with a diverted discharge ratio of 50% or less. The intake elevation highest above the channel bed diverted the least sediment. The interrelationship between Diverted Discharge Ratio (DDR), Diverted Froude number Ratio (DFrR) and Diverted Sediment load Ratio (SDR) was established in the study. It is recommended that RoR schemes have sand traps downstream of the diversion structures and that turbines are coated with HVOF to overcome the power loss arising due to the excessive erosion of hydro turbines. / AFRIKAANSE OPSOMMING: Sedimentasie en die invloed wat dit het op turbines was die primêre probleem geïdentifiseer in “run-of-river” (RoR) hidrokrag-skemas. Die sediment wat saam met die water uit ‟n rivier uitgekeer word beskadig die inlaatrooster en verlaag kragopwekking in die langtermyn. Skade word aangerig aan alle onderwatertoerusting en masjinerie wat aan sediment blootgestel word. Die grootste probleem tydens die uitkering van water is die growwe sediment wat daarmee deur die onttrekking inlaat gaan en turbineskade veroorsaak. Soos wat die sedimentlading in die rivier drasties toeneem, sal afslyting en kavitasie van turbinelemme meer gereeld voorkom. Dié probleem kan op twee maniere beperk word. Die een is die gebruik van bestaande opgaardamme stroomop, en die tweede is deur die ondersoek van drie kenmerke van rivierdraaie en uitkeringstrukture, bv. die optimale uitskurings posisie in 'n rivierdraai (sonder ʼn struktuur) om die diepste uitskuringposisie op die buitekant van die draai te bepaal, die optimale uitkeringsstruktuuroriëntasie wat maksimum uitskuring verseker en die sediment uitkering beperk, en die lading van sedimentonttrekking deur die inlaat. Die eerste doelwit van die navorsing is ondersoek deur ʼn fisiese model te bou van ʼn kronkelkanaal en te bepaal waar die diepste uitskring plaasvind op die buitekant van die draai. Die tweede doelwit van die studie was om die optimale uitkeringshoek te bepaal vir 'n uitkeringstruktuur sodat die uitskuring by die inlaat ʼn maksimum is om die uitkering van sediment te beperk. ʼn Derde doelwit was om die akkuraatheid van ʼn wiskundige model se uitskuring voorspelling te toets teen die waargenome laboratorium resultate. Die finale doelwit was om vir verskillende inlaatontwerpe, rivier- en uitkeervloeie die sedimentladings wat uitgekeer word te ondersoek. Die eerste eksperiment in die kronkelende kanaal was voorberei om die optimale uitskuring in die draai te bepaal. Dit is gevolg deur toetse met uitkeerstrukture by die maksimum uitskurings posisie te plaas en die hoek van die struktuur dan te verander met verwysing na die vloeirigting. Die vloeirigting vektor was as 'n raaklyn geplaas op die kanaal draai en georiënteer met hoeke: 0⁰, 30⁰, 45⁰ en 60⁰, in die rigting van die draai. Die optimale uitskurings posisie was aan die buiterand van die kanaal draai gevind, ongeveer 60⁰ in die draai in. Die maksimum uitskuur posisie van 'n 90⁰ kanaal draai stem ooreen met SC en CHES (1992) se resultaat van 60⁰ in die draai in. Daar was ook genoegsame bewyse dat 'n optimale uitkeerwerke oriëntasie van 30⁰ die doeltreffendste en effektiefste uitskuring sal gee. Numeriese simulasies is deur middel van 'n twee dimensionele wiskundige model CCHE 2D (hidro- en sedimentdinamika) uitgevoer. Die numeriese resultate was vergelyk met die laboratoriumresultate om die CCHE program te verifieer as 'n voordelige simulasie program. Daar is gevind dat die wiskundige model die uitskuurdieptes by die inlate met ʼn akkuraatheid van 43.8 % voorspel, wat binne die akkuraatheid is van die sedimentvervoervergelyking wat deur die numeriese model gebruik word. Die finale eksperiment was die uitkering van sediment met verskillende inlaathoogtes en uitkerings sedimentladings. Uit die toetse was dit duidelik dat 'n lae sediment uitkeerverhouding behaal kan word met 'n uitkeerverhouding van 50% en minder. Verdere waarnemings het ook gewys dat die inlaathoogte van die uitkeerstruktuur met die optimale resultate die hoogste bokant die rivierbedding was. Die verwantskap tussen die uitgekeerde deurstromingverhouding, die uitgekeerde Froude getal verhouding en die uitgekeerde sedimentlading is bepaal in die navorsing. Dit word aanbeveel dat sandvangkanale stroomaf van uitkeerwerke geplaas word en dat turbines met HVOF as bedekkingsmateriaal beskerm word om kragverliese as gevolg van buitensporige erosie van die turbines te voorkom.

本文主要探討存戶決策對銀行投資組合決策之影響，及其可能引發之系統性風險的程度。以代表性銀行開始分析，假設銀行投資於安全性資產與風險性資產，而存戶依其決策可能會產生兩個均衡同時存在之情形。本文利用演化的力量，發展出以風險性資產報酬率作為均衡選擇之標準：當風險性資產報酬率低於某一水準時，擠兌的均衡會單獨發生。並比較銀行考慮擠兌發生可能性時，其投資組合承擔風險之程度，及可能引發的銀行倒閉機率大小。推廣至二家模型，發現愈多家銀行在作投資決策考慮存戶擠兌之可能性，愈會增加經濟體系中之系統性風險，顯示銀行與存戶之衝突與金融體系之脆弱性。最後討論資本適足性管制與央行最後貸款人角色能否有效降低銀行倒閉機率。

Red ochre, an iron-oxide mineral has been found in prehistoric sites worldwide, in many types of sites. It has been noted as being present, but has not been afforded the status of artifact with the same level of analysis as other artifacts such as ceramics or lithics. McCullough’s Run, a multi-component prehistoric cemetery in Indiana contained red ochre with cremation burials. By treating red ochre as more than incidental to the site, but as an artifact, new information about mortuary behavior was learned. The red ochre was from locally or regionally available materials, and was placed with the deceased during cremation. Therefore, red ochre must be treated as an artifact for comparing, contrasting information intraand inter-site. / Red ochre as an artifact -- What exactly is red ochre? -- Setting the stage -- The site : McCullough's Run, Bartholomew County, Indiana, 12B1036 -- Red ochre : the artifact. / Department of Anthropology

A significant proportion of the world's population and physical assets are located in low lying coastal zones. Accurate prediction of wave induced run-up and overtopping of sea defences are important in defining the extent and severity of wave action, and in assessing risk to people and property from severe storms and tsunamis. This thesis describes a one-dimensional numerical model based on the Boussinesq equations of Madsen and Sorensen (1992) and the non-linear shallow water equations. The model is suitable for simulating propagation of weakly non-linear and weakly dispersive waves from intermediate to zero depth, such that any inundation and/or overtopping caused by the incoming waves is also calculated as part of the simulation. Wave breaking is approximated by locally switching to the non-linear shallow water equations, which can model broken waves as bores. A piston paddle wavemaker is incorporated into the model for complete reproduction of laboratory experiments. A domain mapping technique is used in the vicinity of the paddle to transform a time-varying domain into a fixed domain, so that the governing equations can be more readily solved. First, various aspects of the numerical model are verified against known analytical and newly derived semi-analytical solutions. The complete model is then validated with laboratory measurements of run-up and overtopping involving solitary waves. NewWave focused wave groups, which give the expected shape of extreme wave events in a linear random sea, are used for further validation. Simulations of experiments of wave group run-up on a plane beach yield very good agreement with the measured run-up distances and free surface time series. Wave-by-wave overtopping induced by focused wave groups is also successfully simulated with the model, with satisfactory agreement between the experimental and the predicted overtopping volumes. Repeated simulations, now driven by second order paddle displacement signals, give insight into second order error waves spuriously generated by using paddle signals derived from linear theory. Separation of harmonics reveals that the long error wave is significantly affecting the wave group shape and leading to enhanced runu-up distances and overtopping volumes. An extensive parameter study is carried out using the numerical model investigating the influence on wave group run-up of linear wave amplitude at focus, linear focus location, and wave group phase at focus. For a given amplitude, both the phase and the focus location significantly affect the wave group run-up. It is also found that the peak optimised run-up increases with the wave amplitude, but wave breaking becomes an inhibiting factor for larger waves. This methodology is proposed for extreme storm wave induced run-up analysis.

Alors que la démolition tend à s'imposer au sein de l'intervention publique sur le parc HLM des années 1950-1970, ce mode d'action commence aussi à être utilisé en réponse aux problèmes rencontrés par des tours et des barres en copropriété. On peut cependant se demander si les difficultés d'un grand ensemble en copropriété sont assimilables à celles d'un grand ensemble HLM et si la démolition ne prend pas un sens différent dans ce contexte particulier. Pour répondre à ces questions, cette thèse étudie l'une des premières copropriétés des trente glorieuses qui a fait l'objet d'une action publique et de démolitions : Les Bosquets à Clichy-sous-Bois/Montfermeil (93). En explorant l'histoire de cet ensemble immobilier, cette recherche montre que les problèmes qu'il a connus sont propres au statut juridique de la copropriété. Ils sont principalement dus à des difficultés de gestion, liées aux financements publics qui sont à l'origine de sa construction, à la conjoncture immobilière dans laquelle il a été livré et à des malfaçons juridiques. Or, cette dimension propre à la gestion n'a pas été prise en compte au sein de l'intervention publique déployée sur ce site depuis 1981. Celle-ci a transposé à la copropriété les schémas cognitifs et normatifs appliqués aux ensembles HLM de même typologie architecturale et urbaine. Privilégiant la démolition, les restructurations du bâti et la diversification du peuplement, l'action publique a aggravé les déséquilibres de gestion de la copropriété. Le relogement a constitué un nouveau facteur de fragilisation des trajectoires des ménages – en particulier des propriétaires occupants – déjà fortement pénalisés par les difficultés de la copropriété / While demolition of public rental housing projects from the 1960s and 1970s is an obvious public intervention nowadays, we are starting to witness a similar handling of condominium tower blocks. However, one may ask whether similar interventions for these two different housing types lead to different outcomes. To answer these questions, this thesis analyses one of the first condominium project from the Trente Glorieuses period (1945-1975) that was the target of such intervention : Les Bosquets located in Clichy sous Bois/Montfermeil (a Parisian suburb). Looking at the history of that housing complex, this research shows that the problems relate to the legal status of the joint ownership. These problems arose from management difficulties linked to the public financing that has permitted its construction, to the real estate circumstances of its delivery, and to juridical defects. It appears that this management dimension wasn't considered by the public intervention implemented at that site since 1981. Cognitive and normative patterns originally applied to moderate rental estates with similar architectural and urban typology were simply transferred to the condominium. Favouring demolition, redevelopments of the building, and diversification of residents, public interventions have aggravated imbalances in the management of the condominium. Rehousing has also participated in the weakening of the residential itineraries of the households (especially owner occupiers) already strongly penalised by the condominium's difficulties