John Ewart Hardern Shaw

John Ewart Hardern Shaw
Coventry University | CU · Department of Design and Visual Arts

PhD

About

64
Publications
8,004
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
945
Citations
Introduction
Ewart Shaw retired in 2018 from the Department of Statistics, University of Warwick, and then studied Fine Art at Coventry University until hit by long covid. His research has involved Bayesian inference, numerical and graphical methods, spherical and related codes, and medical statistics. His current interests include multimedia, neurodiversity, the tension between pattern and randomness, and uses of mathematics, science and computing in art and music.
Additional affiliations
January 1986 - August 2018
The University of Warwick
Position
  • Lecturer
October 1983 - December 1985
University of Nottingham
Position
  • Research Assistant
September 1975 - September 1983
St Bartholomew's Hospital Medical School
Position
  • Computer programmer/Statistician
Education
September 2018 - June 2021
Coventry University
Field of study
  • Fine Art
October 1983 - March 1988
University of Nottingham
Field of study
  • Aspects of Numerical Integration and Summarisation
September 1974 - September 1975
Imperial College London
Field of study
  • Statistics

Publications

Publications (64)
Article
Full-text available
Practical Bayesian statistics with realistic models usually gives posterior distributions that are analytically intractable, and inferences must be made via numerical integration. In many cases, the integrands can be transformed into periodic functions on the unit $d$-dimensional cube, for which quasirandom sequences are known to give efficient num...
Article
Full-text available
Objective To investigate whether surgery by open reduction and internal fixation provides benefit compared with non-operative treatment for displaced, intra-articular calcaneal fractures. Design Pragmatic, multicentre, two arm, parallel group, assessor blinded randomised controlled trial (UK Heel Fracture Trial). Setting 22 tertiary referral hospit...
Preprint
Full-text available
The Dirichlet Linkage, defined here, gives a simple way to explain and visualise the Riemann Hypothesis. It also suggests interesting possibilities for artwork. I produced this document after attending an Open Day at Coventry University, and rashly agreeing that I might indeed use skills from my statistical career during my proposed retirement pla...
Article
Full-text available
The accelerated failure time (AFT) models have proved useful in many contexts, though heavy censoring (as for example in cancer survival) and high dimensionality (as for example in microarray data) cause difficulties for model fitting and model selection. We propose new approaches to variable selection for censored data, based on AFT models optimiz...
Article
Full-text available
In public health, demography and sociology, large-scale surveys often follow a hierarchical data structure as the surveys are based on multistage stratified cluster sampling. The appropriate approach to analyzing such survey data is therefore based on nested sources of variability which come from different levels of the hierarchy. When the variance...
Preprint
Full-text available
This essay considers the creative process, 'left-brain' v. 'right-brain' thinking, neurodiversity, order/chaos, pattern/randomness, heart-line/head-line and palmistry. It was a blueprint for what would have been my final year fine art dissertation.
Preprint
Full-text available
This outline of how Gerhard Richter has been an influence on my artwork is extracted from an essay written while at Coventry University. Most importantly, it shows (pages 6-11) remarkable similarities between certain of Richter's paintings and my own abstract photographs, which were taken long after visiting a Richter exhibition (1989) but before I...
Article
Full-text available
Background: Birth control is the conscious control of the birth rate by methods which temporarily prevent conception by interfering with the normal process of ovulation, fertilization, and implantation. High contraceptive prevalence rate is always expected for controlling births for those countries that are experiencing high population growth rate....
Article
Full-text available
We develop variable selection approaches for accelerated failure time models, consisting of a group of algorithms based on a synthesis of two widely used techniques in the area of variable selection for survival analysis—the Buckley–James method and the Dantzig selector. Two algorithms are based on proposed modified Buckley–James estimating methods...
Research
Full-text available
This package impute the last largest censored observation(s) under weighted least squares and hence brings less bias and more efficient estimates for AFT models.
Research
Full-text available
To compute the original and modified jackknife estimates of Kaplan-Meier estimators.
Research
Full-text available
This provides new approaches to variable selection for AFT model.
Article
Full-text available
For studying or reducing the bias of functionals of the Kaplan-Meier survival estimator, the jackknifing approach of Stute and Wang (1994) is natural. We have studied the behavior of the jackknife estimate of bias under different configurations of the censoring level, sample size, and the censoring and survival time distributions. The empirical res...
Article
Full-text available
When observations are subject to right censoring, weighted least squares with appropriate weights (to adjust for censoring) is sometimes used for parameter estimation. With Stute's weighted least squares method, when the largest observation ($Y_{(n)}^+$) is censored, it is natural to apply the redistribution to the right algorithm of Efron (1967)....
Patent
Description The package is for computing both the original and modified jackknife estimates of Kaplan-Meier estimators. Depends imputeYn License GPL-2 R topics documented: jackknifeKME-package . jackknifeKME-package Jackknife Estimates of Kaplan-Meier Estimators or Integrals Description The package is for computing both the original and modified ja...
Patent
Depends imputeYn, glmnet, lars, quadprog, R (>= 3.0.2) Description Provides new approaches to variable selection for censored data including its high-dimensionality, based on AFT models optimized using regularized weighted least squares. Ap-proaches namely, a weighted elastic net, an adaptive elastic net, and two of their exten-sions by adding cens...
Patent
Description If Kaplan-Meier weights are involved in modeling, especially solving AFT model us-ing weighted least squares method, the package gives alternative choices to the Efron's tail cor-rection approach that brings estimates with less bias and more efficiency.
Article
Full-text available
This study examines the selected determinants of contraceptive prevalence among 10-49 aged ever-married women in Bangladesh and their true impact on the contraception prevalence rate (CPR). It applies a multilevel logistic regression analysis to draw valid conclusions about the effects of the selected determinants on CPR using the 2004 Bangladesh D...
Conference Paper
Full-text available
Links between electrostatics and the zeros of univariate orthogonal polynomials are described, focussing on the Hermite polynomials because of their importance in, for example, asymptotic approximations and numerical integration for Bayesian inference. Extensions to the multivariate case are investigated.
Article
A simple parametrization, built from the definition of cubic splines, is shown to facilitate the implementation and interpretation of penalized spline models, whatever configuration of knots is used. The parametrization is termed value-first derivative parametrization. Inference is Bayesian and explores the natural link between quadratic penalties...
Article
Full-text available
Article
Full-text available
Time to first seizure is a common outcome in antiepileptic drug (AED) studies. Previous studies have typically failed to find statistically significant differences between carbamazepine (CBZ) and valproate (VPS). We re-analyzed a meta-analysis comparing CBZ and VPS monotherapy with new powerful statistical methods that incorporate baseline seizure...
Article
A fundamental problem with the latent-time framework in competing risks is the lack of identifiability of the joint distribution. Given observed covariates along with assumptions as to the form of their effect, then identifiability may obtain. However it is difficult to check any assumptions about form since a more general model may lose identifiab...
Article
In studies of recurrent events, such as epileptic seizures, there can be a large amount of information about a cohort over a period of time, but current methods for these data are often unable to utilize all of the available information. The paper considers data which include post-treatment survival times for individuals experiencing recurring even...
Article
A class of parametric dynamic survival models are explored in which only limited parametric assumptions are made, whilst avoiding the assumption of proportional hazards. Both the log-baseline hazard and covariate effects are modelled by piecewise constant and correlated processes. The method of estimation is to use Markov chain Monte Carlo simulati...
Research
Full-text available
The following relationship is quite pretty but quite useless. 2 Σ_{k=0}^{17} (sin kπ/12 − cos kπ/5 = √ 0 + √ 1 + √ 2 + √ 3 + √ 4 + √ 5 + √ 6.
Article
Much current analysis of cancer registry data uses the semiparametric proportional hazards Cox model. In this paper, the time-dependent effect of various prognostic indicators on breast cancer survival times from the West Midlands Cancer Intelligence Unit are investigated. Using Bayesian methodology and Markov chain Monte Carlo estimation methods,...
Article
Full-text available
This article illustrates how the hypergeometric function H. in the J programming language (www.jsoftware.com) can be used to create cumulative distribution functions for the Normal, beta, gamma, chi-squared, F, binomial, Poisson and t probability distributions.
Technical Report
Full-text available
This is an introduction to the J language (www.jsoftware.com) as of 1999. Much of the description is still relevant and may be useful; it is uploaded to ResearchGate for historical reasons, as I still occasionally receive requests for it, but all my Warwick University email and files have been deleted. This is the most recent version, A4 size, and...
Conference Paper
Full-text available
1 Background The Bayesian approach arises naturally in probabilistic risk analysis, but typically in practice demands the evaluation of many analytically intractable and high-dimensional integrals of the form π(θ)q(θ)dθ, where π(·) is an unnormalised posterior density, θ is a d-dimensional vector of unknowns (typically model parameters), and q(·) i...
Article
Full-text available
This is an introduction to the J language (www.jsoftware.com) as of 1999. Much of the description is still relevant and may be useful; it is uploaded to ResearchGate for historical reasons, as I still occasionally receive requests for it, but all my Warwick University email and files have been deleted. This version is for US letter size
Article
Central to the activities of any industry is the modernization of its assets. This paper describes the application of statistical methods to the asset management plan of a large water company. A model for the underground assets is given in detail, and the extensive computations required are illustrated. A Bayesian model shows the consequences of va...
Article
Full-text available
The original paper by Michael Evans and Tim Swartz appeared in the August 1995 issue of Statistical Science (10 254-272). Due to a publication error the following discussion of the paper was not included in that issue.
Chapter
Numerical integration in Bayesian inference is equivalent to using a discrete approximation to the posterior distribution. I discuss how this can be done efficiently, emphasising the importance of adaptive methods to find a suitable weight function. In high-dimensional problems, quasirandom sequences are recommended for producing a discretised post...
Article
A natural extension of generalised linear models is to use a cubic spline as the underlying regression function. The resulting models are usually analytically intractable, but recent developments in numerical approaches to Bayesian statistics allows the models to be analysed efficiently. The models are illustrated with an example from biological as...
Article
One of the main obstacles to the routine implementation of Bayesian methods has been the absence of efficient algorithms for carrying out the computational tasks implicit in the Bayesian approach. In this paper, recent progress towards overcoming this problem is reviewed. In particular, novel numerical integration and interpolation methods, which e...
Article
A survey of remedial teaching based on a particular method and carried out in an NHS clinic over 2 years has been presented. The findings suggest a differential pattern of improvement both for reading and spelling as well as for type of learning difficulty, whether 'backward' or 'specifically retarded'. A plea for early recognition of specific lear...
Article
Numerical integration techniques now exist which permit a very flexible approach to Bayesian modelling. Low dimensional summaries can be extracted from joint posterior distributions of up to 15 dimensions. The sensitivity of particular summaries to changes in both the model and the prior can thus be investigated. This paper discusses the various as...
Article
Routine implementation of the Bayesian paradigm requires an efficient approach to the calculation and display of posterior or predictive distributions for given likelihood and prior specifi- cations. In this paper we shall review some of the analytic and numerical approaches currently available, describing in detail a numerical integration strategy...
Article
Full-text available
The relationship between free erythrocyte protoporphyrin and conventional indices of iron status was studied in 49 mothers and their infants. Maternal venous blood samples were collected at 34 weeks gestation and at delivery. The corresponding infant blood samples were collected from the umbilical cord and at age 6 weeks. In each case free erythroc...
Article
A retrospective study was made of 93 patients aged 60 years or greater presenting with renal failure between 1976 and 1981 to assess survival and analyse the factors influencing it. Two-year survival of 28 patients with acute renal failure was 50 per cent and of 65 patients with chronic renal failure it was 40 per cent. Higher initial plasma urea a...
Article
Full-text available
S3341, a new hypotensive agent which binds to alpha2-receptors in animal brain preparations, was studied in normal healthy male volunteers. A dose ranging study with 15 and 25 micrograms/kg of S3341 was performed in a double blind, placebo controlled randomised and balanced manner with 3 subjects. A decrease or BP without noticeable sedation (asses...
Article
We investigated the effect of flecainide on systolic time intervals in 6 healthy subjects. A randomized, double blind, cross-over study was designed using flecainide (2 mg/kg), propranolol (0.2 mg/kg) or saline. In comparison with placebo, flecainide increased heart rate, diastolic BP, QT-Index, QS2-Index, PEP, and reduced LVET-Index (p less than 0...
Article
Full-text available
The DC II mouse chondrosarcoma is a potentially valuable radiobiological tumour system since it has been observed to recover from radiation injury by regrowth from clones that may be counted in histological sections. Unfortunately, the normal growth of this tumour following s.c. implantation in the thigh is irregular both in the time before growth...
Article
Blood glucose, lipoproteins, non-esterified fatty acids, fibrinogen and intermediary metabolites were measured in twenty-two diet failed maturity onset diabetics during a double blind crossover study of metformin (1.7 g/day) or clofibrate (2 g/day) therapy. Patients received combined therapy for a final two month period. Four patients had 12 hour m...
Article
Plasma fibrinogen was measured in 285 diabetics (age range 15-85 years) and 209 controls (age range 23-74 years). Plasma fibrinogen concentration showed a positive skew distribution and an approximate normal distribution was obtained by log transformation. The mean log plasma fibrinogen in the diabetics was significantly increased (p less than 0.00...
Article
Well—mainly for statistics. This is a collection of over 2,000 quotations from the famous, (e.g., Hippocrates: 'Declare the past, diagnose the present, foretell the future'), the infamous (e.g., Stalin: 'One death is a tragedy, a million deaths is a statistic'), and the cruelly neglected (modesty forbids...). Such quotes help me to: • appeal to a h...

Projects

Project (1)
Project
Using mathematics to produce artwork: aleatory music; generative art. Analysing and synthesising sound using wavelets. Proving the Riemann hypothesis.