An Ensemble Method for Combining Subset-Specific Algorithm Fits

The Subsemble algorithm is a general subset ensemble prediction method, which can be used for small, moderate, or large datasets. Subsemble partitions the full dataset into subsets of observations, fits a specified underlying algorithm on each subset, and uses a unique form of k-fold cross-validation to output a prediction function that combines the subset-specific fits. An oracle result provides a theoretical performance guarantee for Subsemble. The paper, "Subsemble: An ensemble method for combining subset-specific algorithm fits" is authored by Stephanie Sapp, Mark J. van der Laan & John Canny (2014) .

Using CF-Compliant Calendars with Climate Projection Data

Support for all calendars as specified in the Climate and Forecast (CF) Metadata Conventions for climate and forecasting data. The CF Metadata Conventions is widely used for distributing files with climate observations or projections, including the Coupled Model Intercomparison Project (CMIP) data used by climate change scientists and the Intergovernmental Panel on Climate Change (IPCC). This package specifically allows the user to work with any of the CF-compliant calendars (many of which are not compliant with POSIXt). The CF time coordinate is formally defined in the CF Metadata Conventions document available at < https://cf-convention.github.io/Data/cf-conventions/cf-conventions-1.13/cf-conventions.html#time-coordinate>.

Methods for Fitting Binomial Models with the Generalized Complementary Log-Log Function

Functionality for fitting a flexible link function in binomial regression based on a latent negative-binomial count. The original link function was proposed by Aranda-Ordaz (1981), and motivation via a negative-binomial latent count by van der Veen and Hui (in prep).

Signal Processing

R implementation of the 'Octave' package 'signal', containing a variety of signal processing tools, such as signal generation and measurement, correlation and convolution, filtering, filter design, filter analysis and conversion, power spectrum analysis, system identification, decimation and sample rate change, and windowing.

Bayesian Latent Variable Analysis

Fit a variety of Bayesian latent variable models, including confirmatory factor analysis, structural equation models, and latent growth curve models. References: Merkle & Rosseel (2018) ; Merkle et al. (2021) .

Network Meta-Analysis Using Bayesian Methods

Network meta-analyses (mixed treatment comparisons) in the Bayesian framework using JAGS. Includes methods to assess heterogeneity and inconsistency, and a number of standard visualizations. van Valkenhoef et al. (2012) ; van Valkenhoef et al. (2015) .

Multiple Imputation of Covariates by Substantive Model Compatible Fully Conditional Specification

Implements multiple imputation of missing covariates by Substantive Model Compatible Fully Conditional Specification. This is a modification of the popular FCS/chained equations multiple imputation approach, and allows imputation of missing covariate values from models which are compatible with the user specified substantive model.

"Hit and Run" and "Shake and Bake" for Sampling Uniformly from Convex Shapes

The "Hit and Run" Markov Chain Monte Carlo method for sampling uniformly from convex shapes defined by linear constraints, and the "Shake and Bake" method for sampling from the boundary of such shapes. Includes specialized functions for sampling normalized weights with arbitrary linear constraints. Tervonen, T., van Valkenhoef, G., Basturk, N., and Postmus, D. (2012) . van Valkenhoef, G., Tervonen, T., and Postmus, D. (2014) .

Derivation of Regression-Based Normative Data

Normative data are often used to estimate the relative position of a raw test score in the population. This package allows for deriving regression-based normative data. It includes functions that enable the fitting of regression models for the mean and residual (or variance) structures, test the model assumptions, derive the normative data in the form of normative tables or automatic scoring sheets, and estimate confidence intervals for the norms. This package accompanies the book Van der Elst, W. (2024). Regression-based normative data for psychological assessment. A hands-on approach using R. Springer Nature.

The R Package Ada for Stochastic Boosting

Performs discrete, real, and gentle boost under both exponential and logistic loss on a given data set. The package ada provides a straightforward, well-documented, and broad boosting routine for classification, ideally suited for small to moderate-sized data sets.