Likelihood-based methods for model fitting and assessment, prediction and intervention analysis of count time series following generalized linear models are provided. Models with the identity and with the logarithmic link function are allowed. The conditional distribution can be Poisson or Negative Binomial.

Simulation of random vectors from truncated multivariate normal and t distributions based on the algorithms proposed by Yifang Li and Sujit K. Ghosh (2015) <doi:10.1080/15598608.2014.996690>.

Helps the user to build and register schema descriptions of disorganised (messy) tables. Disorganised tables are tables that are not in a topologically coherent form, where packages such as tidyr could be used for reshaping. The schema description documents the arrangement of input tables and is used to reshape them into a standardised (tidy) output format.

Implementation of the transformation of the Mean Opinion Scores (MOS) to be used before applying the rank based statistical techniques. The method and its necessity is described in: Babak Naderi, Sebastian Möller (2020) <arXiv:2004.11490>.

Extends invariant causal prediction (Peters et al., 2016, <doi:10.1111/rssb.12167>) to generalized linear and transformation models (Hothorn et al., 2018, <doi:10.1111/sjos.12291>). The methodology is described in Kook et al. (2023, <doi:10.1080/01621459.2024.2395588>).

Create additional rows and columns on broom::tidy() output to allow for easier control on categorical parameter estimates.

This package provides an interactive interface to the tfrmt package. Users can import, modify, and export tables and templates with little to no code.

Reuse objects with a long processing time either by storing those in disk or memory. Uses caching to identify R objects (e.g., data frames, plots, etc.) and allows repeated access to those. Created with the specific goal of skipping the waiting time for summary tables obtained from large SQL tables. It is extensible to other uses, such as caching plots in Tabler dashboards to reduce waiting times.

Instance feature calculation and evolutionary instance generation for the traveling salesman problem. Also contains code to "morph" two TSP instances into each other. And the possibility to conveniently run a couple of solvers on TSP instances.

Prebuilt shiny modules containing tools for viewing data, visualizing data, understanding missing and outlier values within your data and performing simple data analysis. This extends teal framework that supports reproducible research and analysis.

This package provides a set of tools for managing time-series data, with a particular emphasis on defining various frequency types such as daily and weekly. It also includes functionality for converting data between different frequencies.

It performs the smoothing approach provided by penalized least squares for univariate and bivariate time series, as proposed by Guerrero (2007) and Gerrero et al. (2017). This allows to estimate the time series trend by controlling the amount of resulting (joint) smoothness. --- Guerrero, V.M (2007) <DOI:10.1016/j.spl.2007.03.006>. Guerrero, V.M; Islas-Camargo, A. and Ramirez-Ramirez, L.L. (2017) <DOI:10.1080/03610926.2015.1133826>.

Define general templates with tags that can be replaced by content depending on arguments and objects to modify the final output of the document.

Time series forecasting faces challenges due to the non-stationarity, nonlinearity, and chaotic nature of the data. Traditional deep learning models like Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) process data sequentially but are inefficient for long sequences. To overcome the limitations of these models, we proposed a transformer-based deep learning architecture utilizing an attention mechanism for parallel processing, enhancing prediction accuracy and efficiency. This paper presents user-friendly code for the implementation of the proposed transformer-based deep learning architecture utilizing an attention mechanism for parallel processing. References: Nayak et al. (2024) <doi:10.1007/s40808-023-01944-7> and Nayak et al. (2024) <doi:10.1016/j.simpa.2024.100716>.

This package provides a unified tidyverse-compatible interface to R's machine learning ecosystem - from data ingestion to model publishing. The tl_read() family reads data from files ('CSV', Excel', Parquet', JSON'), databases ('SQLite', PostgreSQL', MySQL', BigQuery'), and cloud sources ('S3', GitHub', Kaggle'). The tl_model() function wraps established implementations from glmnet', randomForest', xgboost', e1071', rpart', gbm', nnet', cluster', dbscan', and others with consistent function signatures and tidy tibble output. Results flow into unified ggplot2'-based visualization and optional formatted gt tables via the tl_table() family. The underlying algorithms are unchanged; tidylearn simply makes them easier to use together. Access raw model objects via the $fit slot for a supervised method, or $fit$model for an unsupervised one. Methods include random forests Breiman (2001) <doi:10.1023/A:1010933404324>, LASSO regression Tibshirani (1996) <doi:10.1111/j.2517-6161.1996.tb02080.x>, elastic net Zou and Hastie (2005) <doi:10.1111/j.1467-9868.2005.00503.x>, support vector machines Cortes and Vapnik (1995) <doi:10.1007/BF00994018>, and gradient boosting Friedman (2001) <doi:10.1214/aos/1013203451>.

The data that is generated from independent and consecutive GillespieSSA runs for a generic biochemical network is formatted as rows and constitutes an observation. The first column of each row is the computed timestep for each run. Subsequent columns are used for the number of molecules of each participating molecular species or "metabolite" of a generic biochemical network. In this way TemporalGSSA', is a wrapper for the R-package GillespieSSA'. The number of observations must be at least 30. This will generate data that is statistically significant. TemporalGSSA', transforms this raw data into a simulation time-dependent and metabolite-specific trial. Each such trial is defined as a set of linear models (n >= 30) between a timestep and number of molecules for a metabolite. Each linear model is characterized by coefficients such as the slope, arbitrary constant, etc. The user must enter an integer from 1-4. These specify the statistical modality utilized to compute a representative timestep (mean, median, random, all). These arguments are mandatory and will be checked. Whilst, the numeric indicator "0" indicates suitability, "1" prompts the user to revise and re-enter their data. An optional logical argument controls the output to the console with the default being "TRUE" (curtailed) whilst "FALSE" (verbose). The coefficients of each linear model are averaged (mean slope, mean constant) and are incorporated into a metabolite-specific linear regression model as the dependent variable. The independent variable is the representative timestep chosen previously. The generated data is the imputed molecule number for an in silico experiment with (n >=30) observations. These steps can be replicated with multiple set of observations. The generated "technical replicates" can be statistically evaluated (mean, standard deviation) and will constitute simulation time-dependent molecules for each metabolite. For SSA-generated datasets with varying simulation times TemporalGSSA will generate a simulation time-dependent trajectory for each metabolite of the biochemical network under study. The relevant publication with the mathematical derivation of the algorithm is (2022, Journal of Bioinformatics and Computational Biology) <doi:10.1142/S0219720022500184>. The algorithm has been deployed in the following publications (2021, Heliyon) <doi:10.1016/j.heliyon.2021.e07466> and (2016, Journal of Theoretical Biology) <doi:10.1016/j.jtbi.2016.07.002>.

Execution of various time series models and choosing the best one either by a specific error metric or by picking the best one by majority vote. The models are based on the "forecast" package, written by Prof. Rob Hyndman.

Enables users to build ToxPi prioritization models and provides functionality within the grid framework for plotting ToxPi graphs. toxpiR allows for more customization than the ToxPi GUI (<https://toxpi.github.io/>) and integration into existing workflows for greater ease-of-use, reproducibility, and transparency. toxpiR package behaves nearly identically to the GUI; the package documentation includes notes about all differences. The vignettes download example files from <https://github.com/ToxPi/ToxPi-example-files>.

Download taxonomic databases, convert them into SQLite format, and query them locally for fast, reliable, and reproducible access to taxonomic data.

Computes distances between taxonomic hierarchy nodes using lineage data retrieved from The Taxonomicon <http://taxonomicon.taxonomy.nl>. For distinct nodes, distance is defined as the reciprocal of the depth of their most recent common ancestor; identical nodes have distance zero. This definition yields an ultrametric within each connected hierarchy. Functions are provided for auditable name resolution, online or user-supplied lineage analysis, clade membership, pairwise and matrix distance calculation, hierarchical clustering, principal coordinates analysis, portable JSON analysis bundles, and cache management. Distance matrices are returned as base R dist objects. The distances represent classification depth rather than evolutionary time or phylogenetic branch length.

Built on top of the tibble package, tibbletime is an extension that allows for the creation of time aware tibbles. Some immediate advantages of this include: the ability to perform time-based subsetting on tibbles, quickly summarising and aggregating results by time periods, and creating columns that can be used as dplyr time-based groups.

This package provides tools for evaluating the trustworthiness of machine learning models in production and research settings. Computes a Stability Index that quantifies the consistency of model predictions across multiple runs or resamples, and a Robustness Score that measures model resilience under small input perturbations. Designed for data scientists, ML engineers, and researchers who need to monitor and ensure model reliability, reproducibility, and deployment readiness.

This package provides a wrapper for the TexTra API <https://mt-auto-minhon-mlt.ucri.jgn-x.jp/>, a web service for translating texts between different languages. TexTra API account is required to use the service.

The main goal of the R package treeDbalance is to provide functions for the computation of several measurements of 3D node imbalance and their respective 3D tree imbalance indices, as well as to introduce the new phylo3D format for rooted 3D tree objects. Moreover, it encompasses an example dataset of 3D models of 63 beans in phylo3D format. Please note that this R package was developed alongside the project described in the manuscript Measuring 3D tree imbalance of plant models using graph-theoretical approaches by M. Fischer, S. Kersting, and L. Kühn (2023) <doi:10.48550/arXiv.2307.14537>, which provides precise mathematical definitions of the measurements. Furthermore, the package contains several helpful functions, for example, some auxiliary functions for computing the ancestors, descendants, and depths of the nodes, which ensures that the computations can be done in linear time, or functions that convert existing formats of 3D tree models of other software into the phylo3D format. Moreover, it comprises functions to extract the graph-theoretical topology without vertices of in- and out-degree 1 of rooted 3D trees as well as to adapt node enumerations to the common phylo format. Most functions of treeDbalance require as input a rooted tree in the phylo3D format, an extended phylo format (as introduced in the R package ape 1.9 in November 2006). Such a phylo3D object must have at least two new attributes next to those required by the phylo format: node.coord', the coordinates of the nodes, as well as edge.weight', the literal weight or volume of the edges. Optional attributes are edge.diam', the diameter of the edges, and edge.length', the length of the edges. For visualization purposes one can also specify edge.type', which ranges from normal cylinder to bud to leaf, as well as edge.color to change the color of the edge depiction. This project was supported by the joint research project DIG-IT! funded by the European Social Fund (ESF), reference: ESF/14-BM-A55-0017/19, and the Ministry of Education, Science and Culture of Mecklenburg-Western Pomerania, Germany, as well as by the project ArtIGROW, which is a part of the WIR!-Alliance ArtIFARM â Artificial Intelligence in Farming funded by the German Federal Ministry of Education and Research (FKZ: 03WIR4805).