Flexible Estimation of Conditional Survival Functions Using Machine Learning

Tools for flexible estimation of conditional survival functions using off-the-shelf machine learning tools. Implements both global and local survival stacking. See Wolock CJ, Gilbert PB, Simon N, and Carone M (2024) .

Partial Orders and Relations

Finds equivalence classes corresponding to a symmetric relation or undirected graph. Finds total order consistent with partial order or directed graph (so-called topological sort).

Aster Models

Aster models (Geyer, Wagenius, and Shaw, 2007, ; Shaw, Geyer, Wagenius, Hangelbroek, and Etterson, 2008, ; Geyer, Ridley, Latta, Etterson, and Shaw, 2013, ) are exponential family regression models for life history analysis. They are like generalized linear models except that elements of the response vector can have different families (e. g., some Bernoulli, some Poisson, some zero-truncated Poisson, some normal) and can be dependent, the dependence indicated by a graphical structure. Discrete time survival analysis, life table analysis, zero-inflated Poisson regression, and generalized linear models that are exponential family (e. g., logistic regression and Poisson regression with log link) are special cases. Main use is for data in which there is survival over discrete time periods and there is additional data about what happens conditional on survival (e. g., number of offspring). Uses the exponential family canonical parameterization (aster transform of usual parameterization). There are also random effects versions of these models.

Advanced 'tryCatch()' and 'try()' Functions

Advanced tryCatch() and try() functions for better error handling (logging, stack trace with source code references and support for post-mortem analysis via dump files).

K-Means with Simultaneous Outlier Detection

An implementation of the 'k-means--' algorithm proposed by Chawla and Gionis, 2013 in their paper, "k-means-- : A unified approach to clustering and outlier detection. SIAM International Conference on Data Mining (SDM13)", and using 'ordering' described by Howe, 2013 in the thesis, Clustering and anomaly detection in tropical cyclones". Useful for creating (potentially) tighter clusters than standard k-means and simultaneously finding outliers inexpensively in multidimensional space.

Continuous Time SEM - 'OpenMx' Based Functions

Original 'ctsem' (continuous time structural equation modelling) functionality, based on the 'OpenMx' software, as described in Driver, Oud, Voelkle (2017) , with updated details in vignette. Combines stochastic differential equations representing latent processes with structural equation measurement models. These functions were split off from the main package of 'ctsem', as the main package uses the 'rstan' package as a backend now -- offering estimation options from max likelihood to Bayesian. There are nevertheless use cases for the wide format SEM style approach as offered here, particularly when there are no individual differences in observation timing and the number of individuals is large. For the main 'ctsem' package, see < https://cran.r-project.org/package=ctsem>.

Inference and Prediction of Generic Physiologically-Based Kinetic Models

Fit and simulate any kind of physiologically-based kinetic ('PBK') models whatever the number of compartments. Moreover, it allows to account for any link between pairs of compartments, as well as any link of each of the compartments with the external medium. Such generic PBK models have today applications in pharmacology (PBPK models) to describe drug effects, in toxicology and ecotoxicology (PBTK models) to describe chemical substance effects. In case of exposure to a parent compound (drug or chemical) the 'rPBK' package allows to consider metabolites, whatever their number and their phase (I, II, ...). Last but not least, package 'rPBK' can also be used for dynamic flux balance analysis (dFBA) to deal with metabolic networks. See also Charles et al. (2022) .

Interface to 'Python' Package 'StepMix'

This is an interface for the 'Python' package 'StepMix'. It is a 'Python' package following the scikit-learn API for model-based clustering and generalized mixture modeling (latent class/profile analysis) of continuous and categorical data. 'StepMix' handles missing values through Full Information Maximum Likelihood (FIML) and provides multiple stepwise Expectation-Maximization (EM) estimation methods based on pseudolikelihood theory. Additional features include support for covariates and distal outcomes, various simulation utilities, and non-parametric bootstrapping, which allows inference in semi-supervised and unsupervised settings.

Elevation and GPS Data Visualisation

Simpler processing of digital elevation model and GPS trace data for use with the 'rayshader' package.

Univariate and Multivariate Model-Based Clustering in Group-Specific Functional Subspaces

The funHDDC algorithm allows to cluster functional univariate (Bouveyron and Jacques, 2011, ) or multivariate data (Schmutz et al., 2018) by modeling each group within a specific functional subspace.