This package provides a robust computational framework for analyzing complex multimodal data. Extends existing state-dependent models to account for diverse data streams, addressing challenges such as varying temporal scales and learner characteristics to improve the robustness and interpretability of findings. For methodological details, see Shaffer, Wang, and Ruis (2025) "Transmodal Analysis" <doi:10.18608/jla.2025.8423>.

This package provides a set of tools for descriptive and predictive analysis of time series data. That includes functions for interactive visualization of time series objects and as well utility functions for automation time series forecasting.

The analysis of microarray time series promises a deeper insight into the dynamics of the cellular response following stimulation. A common observation in this type of data is that some genes respond with quick, transient dynamics, while other genes change their expression slowly over time. The existing methods for detecting significant expression dynamics often fail when the expression dynamics show a large heterogeneity. Moreover, these methods often cannot cope with irregular and sparse measurements. The method proposed here is specifically designed for the analysis of perturbation responses. It combines different scores to capture fast and transient dynamics as well as slow expression changes, and performs well in the presence of low replicate numbers and irregular sampling times. The results are given in the form of tables including links to figures showing the expression dynamics of the respective transcript. These allow to quickly recognise the relevance of detection, to identify possible false positives and to discriminate early and late changes in gene expression. An extension of the method allows the analysis of the expression dynamics of functional groups of genes, providing a quick overview of the cellular response. The performance of this package was tested on microarray data derived from lung cancer cells stimulated with epidermal growth factor (EGF). Paper: Albrecht, Marco, et al. (2017)<DOI:10.1186/s12859-016-1440-8>.

Prediction intervals for ARIMA and structural time series models using importance sampling approach with uninformative priors for model parameters, leading to more accurate coverage probabilities in frequentist sense. Instead of sampling the future observations and hidden states of the state space representation of the model, only model parameters are sampled, and the method is based solving the equations corresponding to the conditional coverage probability of the prediction intervals. This makes method relatively fast compared to for example MCMC methods, and standard errors of prediction limits can also be computed straightforwardly.

Topological data analytic methods in machine learning rely on vectorizations of the persistence diagrams that encode persistent homology, as surveyed by Ali &al (2000) <doi:10.48550/arXiv.2212.09703>. Persistent homology can be computed using TDA and ripserr and vectorized using TDAvec'. The Tidymodels package collection modularizes machine learning in R for straightforward extensibility; see Kuhn & Silge (2022, ISBN:978-1-4920-9644-3). These recipe steps and dials tuners make efficient algorithms for computing and vectorizing persistence diagrams available for Tidymodels workflows.

Interface to TensorFlow Probability', a Python library built on TensorFlow that makes it easy to combine probabilistic models and deep learning on modern hardware ('TPU', GPU'). TensorFlow Probability includes a wide selection of probability distributions and bijectors, probabilistic layers, variational inference, Markov chain Monte Carlo, and optimizers such as Nelder-Mead, BFGS, and SGLD.

The United Nations Sustainable Development Goals (SDGs) have become an important guideline for organisations to monitor and plan their contributions to social, economic, and environmental transformations. The text2sdg package is an open-source analysis package that identifies SDGs in text using scientifically developed query systems, opening up the opportunity to monitor any type of text-based data, such as scientific output or corporate publications. For more information see Meier, Mata & Wulff (2025) <doi:10.32614/RJ-2024-005> and Wulff, Meier & Mata (2024) <doi:10.1007/s11625-024-01516-3>.

Tabu search algorithm for binary configurations. A basic version of the algorithm as described by Fouskakis and Draper (2007) <doi:10.1111/j.1751-5823.2002.tb00174.x>.

Allows users to analyze text and classify emotions such as happiness, sadness, anger, fear, and neutrality. It combines text preprocessing, TF-IDF (Term Frequency-Inverse Document Frequency) feature extraction, and Random Forest classification to predict emotions and map them to corresponding emojis for enhanced sentiment visualization.

This package implements an algorithm for generating maps, known as tile maps, in which each region is represented by a single tile of the same shape and size. The algorithm was first proposed in "Generating Tile Maps" by Graham McNeill and Scott Hale (2017) <doi:10.1111/cgf.13200>. Functions allow users to generate, plot, and compare square or hexagon tile maps.

This package provides a reproducible pipeline for deriving two clinically meaningful cutoffs for disease biomarkers using a unified two-stage framework. The package integrates finite mixture modeling with risk prediction using biomarker plus clinical features, followed by decision curve analysis to evaluate clinical utility. Outputs include biomarker density plots, risk calibration curves, decision curves, and summary tables of diagnostic performance. Designed for researchers in bio-statistics, neurology, and data science, this package emphasizes reproducibility, transparency, and clear clinical relevance.

Fit a threshold regression model for Interval Censored Data based on the first-hitting-time of a boundary by the sample path of a Wiener diffusion process. The threshold regression methodology is well suited to applications involving survival and time-to-event data.

Interfaces with the Hugging Face tokenizers library to provide implementations of today's most used tokenizers such as the Byte-Pair Encoding algorithm <https://huggingface.co/docs/tokenizers/index>. It's extremely fast for both training new vocabularies and tokenizing texts.

Node centrality measures for temporal networks. Available measures are temporal degree centrality, temporal closeness centrality and temporal betweenness centrality defined by Kim and Anderson (2012) <doi:10.1103/PhysRevE.85.026107>. Applying the REN algorithm by Hanke and Foraita (2017) <doi:10.1186/s12859-017-1677-x> when calculating the centrality measures keeps the computational running time linear in the number of graph snapshots. Further, all methods can run in parallel up to the number of nodes in the network.

This package provides a dependency-light OAuth 2.0 <https://www.rfc-editor.org/rfc/rfc6749> client supporting the client-credentials and authorization-code grants with token refresh. Built on curl and jsonlite', with base R's socket server for the redirect listener, avoiding heavier HTTP stacks.

This package implements additional operators for computer vision models, including operators necessary for image segmentation and object detection deep learning models.

Common tree-sitter (<https://tree-sitter.github.io/tree-sitter/>) parsing tools for R. It is meant to be used by other packages that specialize in particular languages and file formats.

Operators and functions provided by base R sometimes lack some features found in other programming languages, such as the ability to concatenate strings using + or to repeat strings using *. This package aims at providing such functionality without breaking existing code, i.e., only statements, that would throw errors in pure base R are patched.

Generic methods for parameter tuning of classification algorithms using multiple scoring functions (Muessel et al. (2012), <doi:10.18637/jss.v046.i05>).

In Cox's proportional hazard model, covariates are modeled as linear function and may not be flexible. This package implements additive trend filtering Cox proportional hazards model as proposed in Jiacheng Wu & Daniela Witten (2019) "Flexible and Interpretable Models for Survival Data", Journal of Computational and Graphical Statistics, <DOI:10.1080/10618600.2019.1592758>. The fitted functions are piecewise polynomial with adaptively chosen knots.

Create browsers for reading full texts from a token list format. Information obtained from text analyses (e.g., topic modeling, word scaling) can be used to annotate the texts.

The Common Workflow Language <https://www.commonwl.org/> is an open standard for describing data analysis workflows. This package takes the raw Common Workflow Language workflows encoded in JSON or YAML and turns the workflow elements into tidy data frames or lists. A graph representation for the workflow can be constructed and visualized with the parsed workflow inputs, outputs, and steps. Users can embed the visualizations in their Shiny applications, and export them as HTML files or static images.

Likelihood ratio and maximum likelihood statistics are provided that can be used as alternatives to p-values Colquhoun (2017) <doi:10.1098/rsos.171085>. Arguments can be either p-values or t-statistics. together with degrees of freedom. For the function tTOlr', the argument twoSided has the default twoSided = TRUE'.