Journal of Statistical Software
Not a member yet
1629 research outputs found
Sort by
TrueSkill Through Time: Reliable Initial Skill Estimates and Historical Comparability with Julia, Python, and R
Get PDFKnowing how individual abilities change is essential in a wide range of activities. The most widely used skill estimators in industry and academia (such as Elo and TrueSkill) propagate information in only one direction, from the past to the future, preventing them from obtaining reliable initial estimates and ensuring comparability between estimates distant in time and space. In contrast, the model TrueSkill Through Time (TTT) propagates all historical information throughout a single causal network, providing estimates with low uncertainty at any given time, enabling reliable initial skill estimates, and ensuring historical comparability. Although the TTT model was published more than a decade ago, it was not available until now in the programming languages with the largest communities. Here we offer the first software for Julia, Python, and R, accompanied by a detailed overview for the general public and an in-depth scientific explanation. After illustrating its basic mode of use, we show how to estimate the learning curves of historical players of the Association of Tennis Professionals. Analytical approximation methods and message-passing algorithms allow inference to be solved efficiently using any low-end computer, even in causal networks with millions of nodes and irregular structures
Local Influence Diagnostics for Nonlinear Mixed Models under the Case-Weight Perturbation Scheme in SAS
No full textThe nonlinear mixed model is a popular tool for analyzing continuous longitudinal data. This paper is primarily concerned with gauging the sensitivity of nonlinear mixed models to influential observations through local influence, which assesses the impact of small perturbations of the likelihood function. Unlike when case deletion is used, in local influence the model only needs to be fitted once, making it much more computationally appealing. The methodology is illustrated with two datasets, establishing that the local influence diagnostic can easily be applied to nonlinear mixed models through the NLMIXED procedure in the SAS software as a tool to identify influential individuals
gptools: Scalable Gaussian Process Inference with Stan
Get PDFGaussian processes (GPs) are sophisticated distributions to model functional data. Whilst theoretically appealing, they are computationally cumbersome except for small datasets. We implement two methods for scaling GP inference in Stan: First, a general sparse approximation using a directed acyclic dependency graph; second, a fast, exact method for regularly spaced data modeled by GPs with stationary kernels using the fast Fourier transform. Based on benchmark experiments, we offer guidance for practitioners to decide between different methods and parameterizations. We consider two real-world examples to illustrate the package. The implementation follows Stan's design and exposes performant inference through a familiar interface. Full posterior inference for ten thousand data points is feasible on a laptop in less than 20 seconds. Details on how to get started using the popular interfaces cmdstanpy for Python and cmdstanr for R are provided
Birth-and-Death Processes in Python: The BirDePy Package
Get PDFBirth-and-death processes (BDPs) form a class of continuous-time Markov chains that are particularly suited to describing the changes in the size of a population over time. Population-size-dependent BDPs (PSDBDPs) allow the rate at which a population grows to depend on the current population size. The main purpose of our new Python package BirDePy is to provide easy-to-use functions that allow the parameters of discretely-observed PSDBDPs to be estimated. The package can also be used to estimate parameters of continuously-observed PSDBDPs, simulate sample paths, approximate transition probabilities, and generate forecasts. We describe in detail several methods which have been incorporated into BirDePy to achieve each of these tasks. The usage and effectiveness of the package is demonstrated through a variety of examples of PSDBDPs, as well as case studies involving annual population count data of two endangered bird species
cubble: An R Package for Organizing and Wrangling Multivariate Spatio-Temporal Data
Get PDFMultivariate spatio-temporal data refers to multiple measurements taken across space and time. For many analyses, spatial and time components can be separately studied: for example, to explore the temporal trend of one variable for a single spatial location, or to model the spatial distribution of one variable at a given time. However for some studies, it is important to analyze different aspects of the spatio-temporal data simultaneously, for instance, temporal trends of multiple variables across locations. In order to facilitate the study of different portions or combinations of spatio-temporal data, we introduce a new class, cubble, with a suite of functions enabling easy slicing and dicing on different spatio-temporal components. The proposed cubble class ensures that all the components of the data are easy to access and manipulate while providing flexibility for data analysis. In addition, the cubble package facilitates visual and numerical explorations of the data while easing data wrangling and modelling. The cubble class and the tools implemented in the package are illustrated with examples from climate data analysis
Emulation and History Matching Using the hmer Package
Get PDFModeling complex real-world situations such as infectious diseases, geological phenomena, and biological processes can present a dilemma: the computer model (referred to as a simulator) needs to be complex enough to capture the dynamics of the system, but each increase in complexity increases the evaluation time of such a simulation, making it difficult to obtain an informative description of parameter choices that would be consistent with observed reality. While methods for identifying acceptable matches to real-world observations exist, for example optimization or Markov chain Monte Carlo methods, they may result in non-robust inferences or may be infeasible for computationally intensive simulators. The techniques of emulation and history matching can make such determinations feasible, efficiently identifying regions of parameter space that produce acceptable matches to data while also providing valuable information about the simulator's structure, but the mathematical considerations required to perform emulation can present a barrier for makers and users of such simulators compared to other methods. The hmer package provides an accessible framework for using history matching and emulation on simulator data, leveraging the computational efficiency of the approach while enabling users to easily match to, visualize, and robustly predict from their complex simulators
DoubleML: An Object-Oriented Implementation of Double Machine Learning in R
Get PDFThe R package DoubleML implements the double/debiased machine learning framework of Chernozhukov, Chetverikov, Demirer, Duflo, Hansen, Newey, and Robins (2018). It provides functionalities to estimate parameters in causal models based on machine learning methods. The double machine learning framework consists of three key ingredients: Neyman orthogonality, high-quality machine learning estimation and sample splitting. Estimation of nuisance components can be performed by various state-of-the-art machine learning methods that are available in the mlr3 ecosystem. DoubleML makes it possible to perform inference in a variety of causal models, including partially linear and interactive regression models and their extensions to instrumental variable estimation. The object-oriented implementation of DoubleML enables a high flexibility for the model specification and makes it easily extendable. This paper serves as an introduction to the double machine learning framework and the R package DoubleML. In reproducible code examples with simulated and real data sets, we demonstrate how DoubleML users can perform valid inference based on machine learning methods
The R Package tipsae: Tools for Mapping Proportions and Indicators on the Unit Interval
Get PDFThe tipsae package implements a set of small area estimation tools for mapping proportions and indicators defined on the unit interval. It provides for small area models defined at area level, including the classical beta regression, zero- and/or one-inflated beta and flexible beta ones, possibly accounting for spatial and/or temporal dependency structures. The models, developed within a Bayesian framework, are estimated through Stan language, allowing fast estimation and customized parallel computing. The additional features of the tipsae package, such as diagnostics, visualization and exporting functions as well as variance smoothing and benchmarking functions, improve the user experience through the entire process of estimation, validation and outcome presentation. A shiny application with a user-friendly interface further eases the implementation of Bayesian models for small area analysis
Generalized Plackett-Luce Likelihoods
Get PDFThe hyper2 package provides functionality to work with extensions of the Bradley-Terry probability model such as Plackett-Luce likelihood including team strengths and reified entities (monsters). The package allows one to use relatively natural R idiom to manipulate such likelihood functions. Here, I present a generalization of hyper2 in which multiple entities are constrained to have identical Bradley-Terry strengths. A new S3 class 'hyper3', along with associated methods, is motivated and introduced. Three datasets are analyzed, each analysis furnishing new insight, and each highlighting different capabilities of the package
Weighted scoringRules: Emphasizing Particular Outcomes When Evaluating Probabilistic Forecasts
Get PDFWhen predicting future events, it is common to issue forecasts that are probabilistic, in the form of probability distributions over the range of possible outcomes. Such forecasts can be evaluated using proper scoring rules. Proper scoring rules condense forecast performance into a single numerical value, allowing competing forecasters to be ranked and compared. To facilitate the use of scoring rules in practical applications, the scoringRules package in R provides popular scoring rules for a wide range of forecast distributions. This paper discusses an extension to the scoringRules package that additionally permits the implementation of popular weighted scoring rules. Weighted scoring rules allow particular outcomes to be targeted during forecast evaluation, recognizing that certain outcomes are often of more interest than others when assessing forecast quality. This introduces the potential for very flexible, user-oriented evaluation of probabilistic forecasts. We discuss the theory underlying weighted scoring rules, and describe how they can readily be implemented in practice using scoringRules. Functionality is available for weighted versions of several popular scoring rules, including the logarithmic score, the continuous ranked probability score, and the energy score. Two case studies are presented to demonstrate this, whereby weighted scoring rules are applied to univariate and multivariate probabilistic forecasts in the fields of meteorology and economics