A systematic approach to error isolation in computerized wastewater simulation models

Activated sludge models are used extensively in the study of wastewater treatment processes. While various commercial implementations of these models are available, there are many people who need to code models themselves using the simulation packages available to them, Quality assurance of such models is difficult. While benchmarking problems have been developed and are available, the comparison of simulation data with that of commercial models leads only to the detection, not the isolation of errors. To identify the errors in the code is time-consuming. In this paper, we address the problem by developing a systematic and largely automated approach to the isolation of coding errors. There are three steps: firstly, possible errors are classified according to their place in the model structure and a feature matrix is established for each class of errors. Secondly, an observer is designed to generate residuals, such that each class of errors imposes a subspace, spanned by its feature matrix, on the residuals. Finally. localising the residuals in a subspace isolates coding errors. The algorithm proved capable of rapidly and reliably isolating a variety of single and simultaneous errors in a case study using the ASM 1 activated sludge model. In this paper a newly coded model was verified against a known implementation. The method is also applicable to simultaneous verification of any two independent implementations, hence is useful in commercial model development.

Repeated blood pressure measurements in a sample of Swedish twins: heritabilities and associations with polymorphisms in the renin-angiotensin-aldosterone system Iliadou, Anastasia, Lichtenstein, Paul, Morgenstern, Ralf, Forsberg, Lena, Svensson, Richard, de Faire, Ulf, Martin, Nicholas, Pedersen, Nancy

Background Twin and family studies have shown that genetic effects explain a relatively high amount of the phenotypic variation in blood pressure. However, many studies have not been able to replicate findings of association between specific polymorphisms and diastolic and systolic blood pressure. Methods In a structural equation-modelling framework the authors investigated longitudinal changes in repeated measures of blood pressures in a sample of 298 like-sexed twin pairs from the population-based Swedish Twin Registry. Also examined was the association between blood pressure and polymorphisms in the angiotensin-I converting enzyme and the angiotensin 11 receptor type 1 with the 'Fulker' test Both linkage and association were tested simultaneously revealing whether the polymorphism is a Quantitative Trait Locus (QTL) or in linkage disequilibrium with the QTL. Results Genetic influences explained up to 46% of the phenotypic variance in diastolic and 63% of the phenotypic variance in systolic blood pressure. Genetic influences were stable over time and contributed up to 78% of the phenotypic correlation in both diastolic and systolic blood pressure. Non-shared environmental effects were characterised by time specific influences and little transmission from one time point to the next. There was no significant linkage and association between the polymorphisms and blood pressure. Conclusions There is a considerable genetic stability in both diastolic and systolic blood pressure for a 6-year period of time in adult life. Non-shared environmental influences have a small long-term effect Although associations with the polymorphisms could not be replicated, results should be interpreted with caution due to power considerations. (C) 2002 Lippincott Williams Wilkins.

The validity and heritability of self-report osteoarthritis in an Australian older twin sample

In order to investigate the genetic and environmental antecedents of osteoarthritis (CA), self-report measures of joint pain, stiffness and swelling were obtained from a population-based sample of 1242 twin pairs over 50 years of age. In order to provide validation for these self-report measures, a subsample of 118 twin pairs were examined according to the American College of Rheumatology clinical and radiographic criteria for the classification of osteoarthritis. A variety of statistical methods were employed to identify the model derived from self-report variables which would provide optimal prediction of these standardised assessments, and structural equation modelling was used to determine the relative influences of genetic and environmental influences on the development of osteoarthritis. Significant genetic effects were found to contribute to osteoarthritis of the hands, hips and knees in women, with heritability estimates ranging from 30-46% depending on the site. In addition, the additive genetic effects contributing to osteoarthritis in various parts of the body were confirmed to be the same. Statistically significant familial aggregation of osteoarthritis in men was also observed, but it was not possible to determine whether this was due to genetic or shared environmental effects.

Genetic and environmental contributions to cannabis dependence in a national young adult twin sample

Background. This paper examines genetic and environmental contributions to risk of cannabis dependence. Method. Symptoms of cannabis dependence and measures of social, family and individual risk factors were assessed in a sample of 6265 young adult male and female Australian twins born 1964-1971. Results. Symptoms of cannabis dependence were common: 11(.)0% of sample (15(.)1% of men and 7(.)8% of women) reported two or more symptoms of dependence. Correlates of cannabis dependence included educational attainment, exposure to parental conflict, sexual abuse, major depression, social anxiety and childhood conduct disorder. However, even after control for the effects of these factors, there was evidence of significant genetic effects on risk of cannabis dependence. Standard genetic modelling indicated that 44(.)7% (95% CI = 15-72(.)2) of the variance in liability to cannabis dependence could be accounted for by genetic factors, 20(.)1% (95 CI = 0-43(.)6) could be attributed to shared environment factors and 35(.)3% (95% CI = 26(.)4-45(.)7) could be attributed to non-shared environmental factors. However, while there was no evidence of significant gender differences in the magnitude of genetic and environmental influences, a model which assumed both genetic and shared environmental influences on risks of cannabis dependence among men and shared environmental but no genetic influences among women provided an equally good fit to the data. Conclusions. There was consistent evidence that genetic risk factors are important determinants of risk of cannabis dependence among men. However, it remains uncertain whether there are genetic influences on liability to cannabis dependence among women.

Model specification tests in nonparametric stochastic regression models

In this paper, we consider testing for additivity in a class of nonparametric stochastic regression models. Two test statistics are constructed and their asymptotic distributions are established. We also conduct a small sample study for one of the test statistics through a simulated example. (C) 2002 Elsevier Science (USA).

Empirical evidence for ultrametric structure in multi-layer perceptron error surfaces

Combinatorial optimization problems share an interesting property with spin glass systems in that their state spaces can exhibit ultrametric structure. We use sampling methods to analyse the error surfaces of feedforward multi-layer perceptron neural networks learning encoder problems. The third order statistics of these points of attraction are examined and found to be arranged in a highly ultrametric way. This is a unique result for a finite, continuous parameter space. The implications of this result are discussed.

The impact of genotyping error on haplotype reconstruction and frequency estimation

The choice of genotyping families vs unrelated individuals is a critical factor in any large-scale linkage disequilibrium (LD) study. The use of unrelated individuals for such studies is promising, but in contrast to family designs, unrelated samples do not facilitate detection of genotyping errors, which have been shown to be of great importance for LD and linkage studies and may be even more important in genotyping collaborations across laboratories. Here we employ some of the most commonly-used analysis methods to examine the relative accuracy of haplotype estimation using families vs unrelateds in the presence of genotyping error. The results suggest that even slight amounts of genotyping error can significantly decrease haplotype frequency and reconstruction accuracy, that the ability to detect such errors in large families is essential when the number/complexity of haplotypes is high (low LD/common alleles). In contrast, in situations of low haplotype complexity (high LD and/or many rare alleles) unrelated individuals offer such a high degree of accuracy that there is little reason for less efficient family designs. Moreover, parent-child trios, which comprise the most popular family design and the most efficient in terms of the number of founder chromosomes per genotype but which contain little information for error detection, offer little or no gain over unrelated samples in nearly all cases, and thus do not seem a useful sampling compromise between unrelated individuals and large families. The implications of these results are discussed in the context of large-scale LD mapping projects such as the proposed genome-wide haplotype map.

On the overabundance of light rare earth elements in terrestrial zircons and its implication for Earth's earliest magmatic differentiation

We present whole-rock and zircon rare earth element (REE) data from two early Archaean gneisses (3.81 Ga and 3.64 Ga) from the Itsaq gneiss complex, south-west Greenland. Both gneisses represent extremely rare examples of unaltered, fresh and relatively undeformed igneous rocks of such antiquity. Cathodoluminescence imaging of their zircons indicates a single crystallisation episode with no evidence for either later metamorphic and/or anatectic reworking or inheritance of earlier grains. Uniform, single-population U/Pb age data confirm the structural simplicity of these zircons. One sample, a 3.64 Ga granodioritic gneiss from the Gothabsfjord, yields a chondrite-normalised REE pattern with a positive slope from La to Lu as well as substantial positive Ce and slight negative Eu anomalies, features generally considered to be typical of igneous zircon. In contrast, the second sample, a 3.81 Ga tonalite from south of the Isua Greenstone Belt, has variable but generally much higher light REE abundances, with similar middle to heavy REE. Calculation of zircon/melt distribution coefficients (D-REE(zircon/melt)) from each sample yields markedly different values for the trivalent REE (i.e. Ce and Eu omitted) and simple application of one set of D-REE(zircon/melt) to model the melt composition for the other sample yields concentrations that are in error by up to two orders of magnitude for the light REE (La-Nd). The observed light REE overabundance in the 3.81 Ga tonalite is a commonly observed feature in terrestrial zircons for which a number of explanations ranging from lattice strain to disequilibrium crystallisation have been proposed and are further investigated herein. Regardless of the cause of light REE overabundance, our study shows that simple application of zircon/melt distribution coefficients is not an unambiguous method for ascertaining original melt composition. In this context, recent studies that use REE data to claim that > 4.3 Ga Hadean detrital zircons originally crystallised from an evolved magma, in turn suggesting the operation of geological processes in the early Earth analogous to those of the present day (e.g. subduction and melting of hydrated oceanic crust), must be regarded with caution. Indeed, comparison of terrestrial Hadean and > 3.9 Ga lunar highland zircons shows remarkable similarities in the light REE, even though subduction processes that have been used to explain the terrestrial zircons have never operated on the Moon. (C) 2002 Elsevier Science B.V. All rights reserved.

Evaluation of various pre-treatments for the dehydration of banana and selection of suitable drying models

The thin-layer drying behaviour of bananas in a beat pump dehumidifier dryer was examined. Four pre-treatments (blanching, chilling, freezing and combined blanching and freezing) were applied to the bananas, which were dried at 50 degreesC with an air velocity of 3.1 m s(-1) and with the relative humidity of the inlet air of 10-35%. Three drying models, the simple model, the two-term exponential model and the Page model were examined. All models were evaluated using three statistical measures, correlation coefficient, root means square error, and mean absolute percent error. Moisture diffusivity was calculated based on the diffusion equation for an infinite cylindrical shape using the slope method. The rate of drying was higher for the pre-treatments involving freezing. The sample which was blanched only did not show any improvement in drying rate. In fact, a longer drying time resulted due to water absorption during blanching. There was no change in the rate for the chilled sample compared with the control. While all models closely fitted the drying data, the simple model showed greatest deviation from the experimental results. The two-term exponential model was found to be the best model for describing the drying curves of bananas because its parameters represent better the physical characteristics of the drying process. Moisture diffusivities of bananas were in the range 4.3-13.2 x 10(-10) m(2)s(-1). (C) 2002 Published by Elsevier Science Ltd.