Open-source Statistical Software Research Articles

2268 - Estimating the Determinants of Cycling: From Area to Road Network Levels

Background It is important to model the determinants of mode choice for many health-related reasons: to find-out what is holding people back from healthy active travel choices, to understand the drivers of detrimental car dependency, and to identify the causal factors over which public authorities have some control. Another motivation is that model results could feed into forecasts of behaviour change following specific interventions. Numerous papers have explored the issue but few have been able to provide results that can feed-into models forecasting uptake following local interventions for a number of reasons including small sample size, lack of a temporal component, and use of explanatory variables that cannot easily be applied beyond the case study region. This gap in the literature was identified and tackled as part of a UK Department for Transport funded project to develop a Cycling Infrastructure Prioritisation Toolkit (CyIPT). Methods Origin-destination (OD) data on commuter mode choice was obtained from the 2001 and 2011 Census. The data was processed to ensure that the 2001 data, which was provided at a different geographic level, as compatible with the 2011 data. It was further processed to create a dependent variable (cycling uptake) and explanatory variables including environmental (e.g. length of off-road cycle infrastructure) and demographic (e.g. age and gender splits) factors, for each desire-line. A model was developed to explain cycling uptake as a function of the explanatory variables and forecast future uptake following simulated interventions. Results Preliminary results show a clear infrastructure signal in the noise surrounding mode choice but other factors seem to be at play. Further research over the coming months will seek to improve the model and our resulting ability to forecast uptake at the local level from specific interventions. Conclusions Newly available data, combined with emerging high performance and open source statistical software, mean that large natural experiments on the impact of new infrastructure on mode choice can now be performed at the national level using geographically detailed data at the origin-destination level. This has the potential to make the prioritisation of new infrastructure much more evidence-based and holds the promise of allowing a-priori benefit-cost estimates to be automated to support transport planners to enable a rapid transition away from fossil fuels and towards healthy travel patterns across the world, in urban and rural areas alike.

Evaluating the portability of satellite derived chlorophyll-a algorithms for temperate inland lakes using airborne hyperspectral imagery and dense surface observations

This study evaluated the performances of twenty-nine algorithms that use satellite-based spectral imager data to derive estimates of chlorophyll-a concentrations that, in turn, can be used as an indicator of the general status of algal cell densities and the potential for a harmful algal bloom (HAB). The performance assessment was based on making relative comparisons between two temperate inland lakes: Harsha Lake (7.99 km2) in Southwest Ohio and Taylorsville Lake (11.88 km2) in central Kentucky. Of interest was identifying algorithm-imager combinations that had high correlation with coincident chlorophyll-a surface observations for both lakes, as this suggests portability for regional HAB monitoring. The spectral data utilized to estimate surface water chlorophyll-a concentrations were derived from the airborne Compact Airborne Spectral Imager (CASI) 1500 hyperspectral imager, that was then used to derive synthetic versions of currently operational satellite-based imagers using spatial resampling and spectral binning. The synthetic data mimics the configurations of spectral imagers on current satellites in earth’s orbit including, WorldView-2/3, Sentinel-2, Landsat-8, Moderate-resolution Imaging Spectroradiometer (MODIS), and Medium Resolution Imaging Spectrometer (MERIS). High correlations were found between the direct measurement and the imagery-estimated chlorophyll-a concentrations at both lakes. The results determined that eleven out of the twenty-nine algorithms were considered portable, with r2 values greater than 0.5 for both lakes. Even though the two lakes are different in terms of background water quality, size and shape, with Taylorsville being generally less impaired, larger, but much narrower throughout, the results support the portability of utilizing a suite of certain algorithms across multiple sensors to detect potential algal blooms through the use of chlorophyll-a as a proxy. Furthermore, the strong performance of the Sentinel-2 algorithms is exceptionally promising, due to the recent launch of the second satellite in the constellation, which will provide higher temporal resolution for temperate inland water bodies. Additionally, scripts were written for the open-source statistical software R that automate much of the spectral data processing steps. This allows for the simultaneous consideration of numerous algorithms across multiple imagers over an expedited time frame for the near real-time monitoring required for detecting algal blooms and mitigating their adverse impacts.

HOW MUCH ARE CHARITY, FUNDRAISING, NGO AND NON-PROFITS CURRENTLY PAYING THEIR NEW STAFF?

This article explores current salary levels and some other related questions using public recruitment data from the CharityJobs website. According to CharityJob, the site is the United Kingdom's busiest one for charity, fundraising, NGO and not for profit jobs. Data collection took place between 4 September and 20 November 2016 with some basic techniques of web scraping (web harvesting or web data extraction), which is a computer software technique of extracting information from websites. All the process is documented at: https://rpubs.com/EduardoWF/charityjobs. The source code in RMarkdown is available for download following GNU General Public License. Everything was prepared with the open-source and freely-accesible statistical computing software R (R version 3.2.0 - http://cran.r-project.org/) and the IDE RStudio (Version 0.99.441 - http://www.rstudio.com/). In addition to presenting these powerful tools and data-exploring techniques, I hope that this article can help the public, specially applicants and workers in civil-society organisations to get an update on salaries and trends in the sector. The jobs analysed here are mostly UK-based ones and published by UK-based organisations. Therefore, the results are not meant to represent the entire sector worldwide. I still hope though that this analysis can provide some positive contribuition to the evolution of work of civil-society organisation in both the southern and the northern hemispheres. This post is based on public data, is my sole responsibility and can in no way be taken to reflect the views of CharityJobs' staff.

Impulse excitation measurement of small changes in elastic moduli and damping using R

ABSTRACTImpulse excitation is commonly used to assess wood stiffness in both scientific and industrial settings. Modes of vibration are excited by a mechanical impulse and their resonant frequencies measured to calculate elastic moduli. It is also possible to assess damping, which is an important property for wood used in musical instruments. This paper covers finer details of the signal analysis necessary when making very accurate measurements. Provided code for Open Source statistical software R is illustrated with data from an experiment tracking small changes in Douglas-fir clear wood during thermal treatment. A frequency domain interpolation method is introduced that enhances the potential precision of the technique to a very high degree, with computational efficiency, even when using inexpensive audio equipment. With multiple measurements, resonant frequencies can be assessed to a precision of 0.1 Hz and better, allowing for tracking of very small changes due to wood treatments or moisture content.

A Computationally Efficient Framework for Automatic Inertial Sensor Calibration

The calibration of (low-cost) inertial sensors has become increasingly important over the past years, since their use has grown exponentially in many applications going from unmanned aerial vehicle navigation to 3-D animation. However, this calibration procedure is often quite problematic since, aside from compensating for deterministic measurement errors due to physical phenomena such as dynamics or temperature, the stochastic signals issued from these sensors in static settings have a complex spectral structure and the methods available to estimate the parameters of these models are either unstable, computationally intensive, and/or statistically inconsistent. This paper presents a new software platform for calibration of the stochastic component in inertial sensor measurement errors based on the generalized method of wavelet moments, which provides a computationally efficient, flexible, user-friendly, and statistically sound tool to estimate and select from a wide range of complex models. In addition, all this is possible also in a robust framework allowing to perform sensor calibration when the data are affected by outliers. The software is developed within the open-source statistical software R and is based on C++ language allowing it to achieve high computational performance.

Semiparametric regression for discrete time-to-event data

Abstract: Time-to-event models are a popular tool to analyse data where the outcome variable is the time to the occurrence of a specific event of interest. Here, we focus on the analysis of time-to-event outcomes that are either intrinsically discrete or grouped versions of continuous event times. In the literature, there exists a variety of regression methods for such data. This tutorial provides an introduction to how these models can be applied using open source statistical software. In particular, we consider semiparametric extensions comprising the use of smooth nonlinear functions and tree-based methods. All methods are illustrated by data on the duration of unemployment of US citizens.

Invited Commentary

Using data from the United Network for Organ Sharing (UNOS) publicly available, deidentified Scientific Registry of Transplant Recipients (SRTR) database, Crawford and colleagues [1Crawford T.C. Magruder J.T. Grimm J.C. et al.A comprehensive risk score to predict prolonged hospital length of stay after heart transplantation.Ann Thorac Surg. 2018; 105: 83-91Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar] performed a multivariate logistic regression analysis to develop a predictive risk score for prolonged postoperative hospital length of stay in a cohort of isolated, adult heart transplantation recipients who survived to discharge from 2003 to 2012. Notable exclusions from their analysis included patients with missing postoperative length of stay data and all patients who died during the postoperative period during their index hospitalization before discharge. Using an approximation of the 90th percentile of length of stays for the entire cohort, the authors defined a prolonged postoperative hospital length of stay as a posttransplant hospital length of stay greater than 30 days. Using the c-statistic to evaluate for discrimination, the authors found their model to be modestly accurate at best (c-statistic 0.61). Multivariate logistic regression has been the cornerstone on which cardiac surgery risk estimation models have been developed over the last 3 decades. The methodology assumes fixed, weighted combinations of independent variables (input) and linearly relates them to the dependent (output) variables under evaluation. However, using logistic regression to model complex postoperative outcomes is insufficient as this method is unable to correctly evaluate the nonlinear and indirect relationships that characterize more complicated clinical scenarios, leading to more often than not inaccurately published risk prediction models. One potential way of more accurately assessing these nonlinear relationships is through the use of advanced machine learning statistical methods. Machine learning techniques, coupled with readily available large electronic datasets and open-source statistical software, can integrate massive feature sets to model many interrelated, nonlinear relationships and higher order interactions. Examples of machine learning methods arriving at more accurate clinical risk prediction models are present throughout the medical literature but are also now beginning to dot the cardiac surgery evidence base as well. Examples include the Bayesian network approach of Loghmanpour and colleagues [2Loghmanpour N.A. Kanwar M.K. Druzdzel M.J. Benza R.L. Murali S. Antaki J.F. A new Bayesian network-based risk stratification model for prediction of short-term and long-term LVAD mortality.ASAIO J. 2015; 61: 313-323Crossref PubMed Scopus (34) Google Scholar] for short- and long-term mortality risk in left ventricular assist device patients using the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) data; and the more recent publication by Allyn and colleagues [3Allyn J. Allou N. Augustin P. et al.A comparison of a machine learning model with EuroSCORE II in predicting mortality after elective cardiac surgery: a decision curve analysis.PLoS One. 2017; 12: e0169772Crossref PubMed Scopus (95) Google Scholar] of an ensemble machine learning risk prediction model that outperformed the European System for Cardiac Operative Risk Evaluation (EuroSCORE) II in a cohort of 6,520 cardiac surgery patients from a single center in France. The idea behind building a risk prediction model with forced linear relationships results in publication of many inaccurate models. Instead, we should strive to assess the true nature of the underlying relationships between clinical-level features through machine learning methodologies in hopes of developing more accurate risk stratification and prediction scores that will help us more precisely inform patient counseling and surgeon practice. A Comprehensive Risk Score to Predict Prolonged Hospital Length of Stay After Heart TransplantationThe Annals of Thoracic SurgeryVol. 105Issue 1PreviewProlonged hospital length of stay (PLOS) after heart transplantation increases cost and morbidity. To better inform care, we developed a risk score to identify patients at risk for PLOS after heart transplantation. Full-Text PDF

Empirically-based modeling and mapping to consider the co-occurrence of ecological receptors and stressors

Part of the ecological risk assessment process involves examining the potential for environmental stressors and ecological receptors to co-occur across a landscape. In this study, we introduce a Bayesian joint modeling framework for use in evaluating and mapping the co-occurrence of stressors and receptors using empirical data, open-source statistical software, and Geographic Information Systems tools and data. To illustrate the approach, we apply the framework to bioassessment data on stream fishes and nutrients collected from a watershed in southwestern Ohio. The results highlighted the joint model's ability to parse and exploit statistical dependencies in order to provide empirical insight into the potential environmental and ecotoxicological interactions influencing co-occurrence. We also demonstrate how probabilistic predictions can be generated and mapped to visualize spatial patterns in co-occurrences. For practitioners, we believe that this data-driven approach to modeling and mapping co-occurrence can lead to more quantitatively transparent and robust assessments of ecological risk.

Spatial uncertainty of a geoid undulation model in Guayaquil, Ecuador

AbstractGeostatistics is a discipline that deals with the statistical analysis of regionalized variables. In this case study, geostatistics is used to estimate geoid undulation in the rural area of Guayaquil town in Ecuador. The geostatistical approach was chosen because the estimation error of prediction map is getting. Open source statistical software R and mainly geoR, gstat and RGeostats libraries were used. Exploratory data analysis (EDA), trend and structural analysis were carried out. An automatic model fitting by Iterative Least Squares and other fitting procedures were employed to fit the variogram. Finally, Kriging using gravity anomaly of Bouguer as external drift and Universal Kriging were used to get a detailed map of geoid undulation. The estimation uncertainty was reached in the interval [-0.5; +0.5] m for errors and a maximum estimation standard deviation of 2 mm in relation with the method of interpolation applied. The error distribution of the geoid undulation map obtained in this study provides a better result than Earth gravitational models publicly available for the study area according the comparison with independent validation points. The main goal of this paper is to confirm the feasibility to use geoid undulations from Global Navigation Satellite Systems and leveling field measurements and geostatistical techniques methods in order to use them in high-accuracy engineering projects.

Breaking walls towards fully open source hydrological modeling

Hydrological models are powerful mathematical tools to address environmental problems and are often used for watershed management and planning. Hydrological models are data driven and the lack of data availability often limits model development. In this paper, we address several challenges in building and running a hydrological model for streamflow simulations based solely on freely available data and open source software. The Soil and Water Assessment Tool (SWAT) hydrological modeling software has been used in the Map Window Geographic Information System (GIS). All spatial and non-spatial data used in this study were obtained from various free of charge online sources. Model calibration and validation represent major challenges following the initial model construction since they involve several trial and error processes to reach acceptable model performances. These critical steps were programmed here as automated scripts in the R open source statistical package. The challenges of model building are described step by step through video tutorials. Using a case study in the Mendoza watershed in Argentina, we show that simulated streamflow exhibits sound agreement with the observed streamflow considering daily time steps (NSE = 0.69, R 2 = 0.72 and Percent bias = +9%). The workflow demonstrated in this study can be applied for other watersheds, especially in data-sparse regions that may lack key regional or local data sets.

A COMPARISON OF ARTIFICIAL NEURAL NETWORKS AND ORDINARY KRIGING DEPTH MAPS OF THE LOWER AND UPPER PANNONIAN STAGE BORDER IN THE BJELOVAR SUBDEPRESSION, NORTHERN CROATIA

Computerised mapping of subsurface strata is possible with a wide range of methods and techniques, such as geostatistical interpolation and stochastic simulations, but also with geomathematical methods. Geomathematical methods are, for example, the use of statistics in geology and the use of artificial neural networks. Artificial neural networks are primarily used in the case of flawed data and data that is in a non-linear relation. The set hypothesis of successful mapping of depth data using this original artificial neural network algorithm is confirmed using statistical analysis and comparison with geostatistical interpolation methods. The algorithm is made in „R“, an open source statistical computing software, and is used on the mapping of depth of the e-log marker „Rs5“ in the Bjelovar Subdepression, Northern Croatia, that is the border between the Lower and Upper Pannonian stages in the Croatian part of the Pannonian Basin System. The neural network architecture that produced the best responses is a network with two hidden layers, with 10 and 6 neurons, respectively. A backpropagation algorithm is used. Two methods were compared by cross-validation and the neural network produced a mean squared error as 16294.5, and Ordinary Kriging produced 14638.35.

Running an open experiment: transparency and reproducibility in soil and ecosystem science

Researchers in soil and ecosystem science, and almost every other field, are being pushed—by funders, journals, governments, and their peers—to increase transparency and reproducibility of their work. A key part of this effort is a move towards open data as a way to fight post-publication data loss, improve data and code quality, enable powerful meta- and cross-disciplinary analyses, and increase trust in, and the efficiency of, publicly-funded research. Many scientists however lack experience in, and may be unsure of the benefits of, making their data and fully-reproducible analyses publicly available. Here we describe a recent ‘open experiment’, in which we documented every aspect of a soil incubation online, making all raw data, scripts, diagnostics, final analyses, and manuscripts available in real time. We found that using tools such as version control, issue tracking, and open-source statistical software improved data integrity, accelerated our team’s communication and productivity, and ensured transparency. There are many avenues to improve scientific reproducibility and data availability, of which is this only one example, and it is not an approach suited for every experiment or situation. Nonetheless, we encourage the communities in our respective fields to consider its advantages, and to lead rather than follow with respect to scientific reproducibility, transparency, and data availability.

MILC: A Microsimulation Model of the Natural History of Lung Cancer

The Microsimulation Lung Cancer (MILC) model was developed to simulate individual trajectories and predict outcomes of lung cancer for populations. The model describes the natural history of lung cancer from a disease-free state to death. Predictions of individual trajectories depend on a set of covariates including age, sex, and smoking behaviors. The module presented here is designed as part of a comprehensive decision-making toolkit for evaluating lung cancer prevention, screening and treatment policies. The MILC package implements the model in the open-source statistical software R. This paper introduces the main components, simulation algorithm, and specifics of the MILC model, validates it by reproducing observed lung cancer incidence trends in the US population, and uses it to make plausible predictions for 50-year-old men and women with a range of smoking histories.

Maximum-likelihood methods for meta-analysis: A tutorial using R

The method of maximum likelihood provides a versatile way to estimate and conduct inference about moderators of effect size in meta-analytic models. The metafor package for the open-source statistical software R offers easy access to this method. We discuss inferential choices that the meta-analyst must make, and advocate the general choice of random-effects methods. We demonstrate the use of the metafor package using data from two meta-analyses that address group processes. These demonstrations illustrate two contrasting approaches to meta-analytic inference: a priori random-effects inference and conditionally random inference. The examples show that these approaches typically lead to the same conclusions when applied correctly.

Longitudinal study of the effects of teat condition on the risk of new intramammary infections in dairy cows

Machine milking–induced alterations of teat tissue may impair local defense mechanisms and increase the risk of new intramammary infections. The objective of the current study was to assess the influence of short-term and long-term alterations of teat tissue and infectious status of the udder quarter on the risk of naturally occurring new intramammary infections, inflammatory responses, and mastitis. Short-term and long-term changes in teat condition of right udder quarters of 135 cows of a commercial dairy farm in Saxony-Anhalt, Germany, were recorded monthly for 10 mo using simple classification schemes. Quarter milk samples were collected from all examined quarters at each farm visit. Bacteriological culture results and somatic cell counts of quarter milk samples were used to determine new inflammatory responses (increase from ≤100,000 cells/mL to >100,000 cells/mL between 2 samples), new infections (detection of a pathogen from a quarter that was free of the same pathogen at the preceding sampling), and new mastitis (combination of new inflammatory response and new infection). Separate Poisson mixed models for new inflammatory responses, new infections, and new mastitis caused by specific pathogens or groups of pathogens (contagious, environmental, major, minor, or any) were used to estimate risk ratios and 95% confidence intervals. Data preparation and parameter estimation were performed using the open source statistical analysis software R. We observed no effect of any variable describing teat condition on the risk of new intramammary infections, inflammatory responses, or mastitis. Intramammary infections of the same udder quarter in the preceding month did not affect risk either.

Using decision trees to manage hospital readmission risk for acute myocardial infarction, heart failure, and pneumonia.

To improve healthcare quality and reduce costs, the Affordable Care Act places hospitals at financial risk for excessive readmissions associated with acute myocardial infarction (AMI), heart failure (HF), and pneumonia (PN). Although predictive analytics is increasingly looked to as a means for measuring, comparing, and managing this risk, many modeling tools require data inputs that are not readily available and/or additional resources to yield actionable information. This article demonstrates how hospitals and clinicians can use their own structured discharge data to create decision trees that produce highly transparent, clinically relevant decision rules for better managing readmission risk associated with AMI, HF, and PN. For illustrative purposes, basic decision trees are trained and tested using publically available data from the California State Inpatient Databases and an open-source statistical package. As expected, these simple models perform less well than other more sophisticated tools, with areas under the receiver operating characteristic (ROC) curve (or AUC) of 0.612, 0.583, and 0.650, respectively, but achieve a lift of at least 1.5 or greater for higher-risk patients with any of the three conditions. More importantly, they are shown to offer substantial advantages in terms of transparency and interpretability, comprehensiveness, and adaptability. By enabling hospitals and clinicians to identify important factors associated with readmissions, target subgroups of patients at both high and low risk, and design and implement interventions that are appropriate to the risk levels observed, decision trees serve as an ideal application for addressing the challenge of reducing hospital readmissions.

Detecting qualitative interactions in clinical trials with binary responses.

This study considers the detection of treatment-by-subset interactions in a stratified, randomised clinical trial with a binary-response variable. The focus lies on the detection of qualitative interactions. In addition, the presented method is useful more generally, as it can assess the inconsistency of the treatment effects among strata by using an a priori-defined inconsistency margin. The methodology presented is based on the construction of ratios of treatment effects. In addition to multiplicity-adjusted p-values, simultaneous confidence intervals are recommended to use in detecting the source and the amount of a potential qualitative interaction. The proposed method is demonstrated on a multi-regional trial using the open-source statistical software R.

Automated analysis of dynamic Ca2+ signals in image sequences.

Intracellular Ca(2+) signals are commonly studied with fluorescent Ca(2+) indicator dyes and microscopy techniques. However, quantitative analysis of Ca(2+) imaging data is time consuming and subject to bias. Automated signal analysis algorithms based on region of interest (ROI) detection have been implemented for one-dimensional line scan measurements, but there is no current algorithm which integrates optimized identification and analysis of ROIs in two-dimensional image sequences. Here an algorithm for rapid acquisition and analysis of ROIs in image sequences is described. It utilizes ellipses fit to noise filtered signals in order to determine optimal ROI placement, and computes Ca(2+) signal parameters of amplitude, duration and spatial spread. This algorithm was implemented as a freely available plugin for ImageJ (NIH) software. Together with analysis scripts written for the open source statistical processing software R, this approach provides a high-capacity pipeline for performing quick statistical analysis of experimental output. The authors suggest that use of this analysis protocol will lead to a more complete and unbiased characterization of physiologic Ca(2+) signaling.

Automated Analysis of Dynamic Ca<sup>2+</sup> Signals in Image Sequences

Intracellular Ca2+ signals are commonly studied with fluorescent Ca2+ indicator dyes and microscopy techniques. However, quantitative analysis of Ca2+ imaging data is time consuming and subject to bias. Automated signal analysis algorithms based on region of interest (ROI) detection have been implemented for one-dimensional line scan measurements, but there is no current algorithm which integrates optimized identification and analysis of ROIs in two-dimensional image sequences. Here an algorithm for rapid acquisition and analysis of ROIs in image sequences is described. It utilizes ellipses fit to noise filtered signals in order to determine optimal ROI placement, and computes Ca2+ signal parameters of amplitude, duration and spatial spread. This algorithm was implemented as a freely available plugin for ImageJ (NIH) software. Together with analysis scripts written for the open source statistical processing software R, this approach provides a high-capacity pipeline for performing quick statistical analysis of experimental output. The authors suggest that use of this analysis protocol will lead to a more complete and unbiased characterization of physiologic Ca2+ signaling.

Detection of Outbreak Signals Using R

Regression analysis is used to fit a periodic model to weekly disease counts for reportable diseases in Missouri. These trigonometric models are then used to obtain upper control limits for the number of cases that would lead to an outbreak signal. The methods, including graphics, are implemented in the open source statistical package R.