Handles univariate non-parametric density estimation with parametric starts and asymmetric kernels in a simple and flexible way. Kernel density estimation with parametric starts involves fitting a parametric density to the data before making a correction with kernel density estimation, see Hjort & Glad (1995) <doi:10.1214/aos/1176324627>. Asymmetric kernels make kernel density estimation more efficient on bounded intervals such as (0, 1) and the positive half-line. Supported asymmetric kernels are the gamma kernel of Chen (2000) <doi:10.1023/A:1004165218295>, the beta kernel of Chen (1999) <doi:10.1016/S0167-9473(99)00010-9>, and the copula kernel of Jones & Henderson (2007) <doi:10.1093/biomet/asm068>. User-supplied kernels, parametric starts, and bandwidths are supported.

This package provides a collection of functions for analyzing data typically collected or used by behavioral scientists. Examples of the functions include a function that compares groups in a factorial experimental design, a function that conducts two-way analysis of variance (ANOVA), and a function that cleans a data set generated by Qualtrics surveys. Some of the functions will require installing additional package(s). Such packages and other references are cited within the section describing the relevant functions. Many functions in this package rely heavily on these two popular R packages: Dowle et al. (2021) <https://CRAN.R-project.org/package=data.table>. Wickham et al. (2021) <https://CRAN.R-project.org/package=ggplot2>.

The kernel ridge regression and the gradient matching algorithm proposed in Niu et al. (2016) <https://proceedings.mlr.press/v48/niu16.html> and the warping algorithm proposed in Niu et al. (2017) <DOI:10.1007/s00180-017-0753-z> are implemented for parameter inference in differential equations. Four schemes are provided for improving parameter estimation in odes by using the odes regularisation and warping.

This package provides wind energy practitioners with an effective machine learning-based tool that estimates a multivariate power curve and predicts the wind power output for a specific environmental condition.

This package contains functions to compute p-values for the one-sample and two-sample Kolmogorov-Smirnov (KS) tests and the two-sample Kuiper test for any fixed critical level and arbitrary (possibly very large) sample sizes. For the one-sample KS test, this package implements a novel, accurate and efficient method named Exact-KS-FFT, which allows the pre-specified cumulative distribution function under the null hypothesis to be continuous, purely discrete or mixed. In the two-sample case, it is assumed that both samples come from an unspecified (unknown) continuous, purely discrete or mixed distribution, i.e. ties (repeated observations) are allowed, and exact p-values of the KS and the Kuiper tests are computed. Note, the two-sample Kuiper test is often used when data samples are on the line or on the circle (circular data). To cite this package in publication: (for the use of the one-sample KS test) Dimitrina S. Dimitrova, Vladimir K. Kaishev, and Senren Tan. Computing the Kolmogorov-Smirnov Distribution When the Underlying CDF is Purely Discrete, Mixed, or Continuous. Journal of Statistical Software. 2020; 95(10): 1--42. <doi:10.18637/jss.v095.i10>. (for the use of the two-sample KS and Kuiper tests) Dimitrina S. Dimitrova, Yun Jia and Vladimir K. Kaishev (2024). The R functions KS2sample and Kuiper2sample: Efficient Exact Calculation of P-values of the Two-sample Kolmogorov-Smirnov and Kuiper Tests. submitted.

This package provides two lightweight keylist S3 classes klist and knlist': extensions of list that enforce unique keys, supporting either mixed named/unnamed elements or fully named elements, ensuring predictable key-value access.

Implementation of Kmeans clustering algorithm and a supervised KNN (K Nearest Neighbors) learning method. It allows users to perform unsupervised clustering and supervised classification on their datasets. Additional features include data normalization, imputation of missing values, and the choice of distance metric. The package also provides functions to determine the optimal number of clusters for Kmeans and the best k-value for KNN: knn_Function(), find_Knn_best_k(), KMEANS_FUNCTION(), and find_Kmeans_best_k().

Density, distribution function, quantile function and random generation for the K-distribution. A plotting function that plots data on Weibull paper and another function to draw additional lines. See results from package in T Lamont-Smith (2018), submitted J. R. Stat. Soc.

Make computer vision tasks approachable in R by leveraging Large Language Models. Providing fine-tuned prompts, boilerplate functions, and input/output helpers for common computer vision workflows, such as classifying and describing images. Functions are designed to take images as input and return structured data, helping users build practical applications with minimal code.

Machine learning, containing several algorithms for supervised and unsupervised classification, in addition to a function that plots the Receiver Operating Characteristic (ROC) and Precision-Recall (PRC) curve graphs, and also a function that returns several metrics used for model evaluation, the latter can be used in ranking results from other packs.

Distance metrics for mixed-type data consisting of continuous, nominal, and ordinal variables. This methodology uses additive and product kernels to calculate similarity functions and metrics, and selects variables relevant to the underlying distance through bandwidth selection via maximum similarity cross-validation. These methods can be used in any distance-based algorithm, such as distance-based clustering. For further details, we refer the reader to Ghashti and Thompson (2024) <doi:10.1007/s00357-024-09493-z> for dkps() methodology, and Ghashti (2024) <doi:10.14288/1.0443975> for dkss() methodology.

Helps make implicit data assumptions explicit by attaching keys to flat-file data that error when those assumptions are violated. Designed for CSV-first workflows without database infrastructure or version control. Provides key definition, assumption checks, join diagnostics, and automatic drift detection via watched data frames that snapshot before each transformation and report cell-level changes.

This package provides a collection of personal helper functions to avoid redundancy in the spirit of the "Don't repeat yourself" principle of software development (<https://en.wikipedia.org/wiki/Don%27t_repeat_yourself>).

This package provides a variable selection procedure, dubbed KKO, for nonparametric additive model with finite-sample false discovery rate control guarantee. The method integrates three key components: knockoffs, subsampling for stability, and random feature mapping for nonparametric function approximation. For more information, see the accompanying paper: Dai, X., Lyu, X., & Li, L. (2021). â Kernel Knockoffs Selection for Nonparametric Additive Modelsâ . arXiv preprint <arXiv:2105.11659>.

This is a C++ mutual information (MI) library based on the k-nearest neighbor (KNN) algorithm. There are three functions provided for computing MI for continuous values, mixed continuous and discrete values, and conditional MI for continuous values. They are based on algorithms by A. Kraskov, et. al. (2004) <doi:10.1103/PhysRevE.69.066138>, BC Ross (2014)<doi:10.1371/journal.pone.0087357>, and A. Tsimpiris (2012) <doi:10.1016/j.eswa.2012.05.014>, respectively.

Allows analyzing time series representing two-dimensional movements. It accepts a data frame with a time (t), horizontal (x) and vertical (y) coordinate as columns, and returns several dynamical properties such as speed, acceleration or curvature.

This package provides tools to calculate the theoretical hydrodynamic response of an aquifer undergoing harmonic straining or pressurization, or analyze measured responses. There are two classes of models here, designed for use with confined aquifers: (1) for sealed wells, based on the model of Kitagawa et al (2011, <doi:10.1029/2010JB007794>), and (2) for open wells, based on the models of Cooper et al (1965, <doi:10.1029/JZ070i016p03915>), Hsieh et al (1987, <doi:10.1029/WR023i010p01824>), Rojstaczer (1988, <doi:10.1029/JB093iB11p13619>), Liu et al (1989, <doi:10.1029/JB094iB07p09453>), and Wang et al (2018, <doi:10.1029/2018WR022793>). Wang's solution is a special exception which allows for leakage out of the aquifer (semi-confined); it is equivalent to Hsieh's model when there is no leakage (the confined case). These models treat strain (or aquifer head) as an input to the physical system, and fluid-pressure (or water height) as the output. The applicable frequency band of these models is characteristic of seismic waves, atmospheric pressure fluctuations, and solid earth tides.

Kernel smoothing for Wishart random matrices described in Daayeb, Khardani and Ouimet (2025) <doi:10.48550/arXiv.2506.08816>, Gaussian and log-Gaussian models using least square or likelihood cross validation criteria for optimal bandwidth selection.

Retrieve data from kintone (<https://www.kintone.com/>) via its API. kintone is an enterprise application platform.

This package provides SAS PROC FORMAT'-like functionality for creating and applying value formats in R. Supports discrete and range-based mapping of values to labels, reverse formatting (invalue), date/time/datetime formatting with built-in SAS format names, multi-label formats, expression labels evaluated at apply-time, case-insensitive matching, import/export of format definitions, and proper handling of missing values (NA, NULL, NaN).

This package contains kernel smoothing tools designed for use by historical dialectologists and philologists for exploring spatial and temporal patterns in noisy historical language data, such as that obtained from historical texts. The main way in which these might differ from other implementations of kernel smoothing is that they assume that the function (linguistic variable) being explored has the form of the relative frequency of a series of discrete possibilities (linguistic variants). This package also offers a way of exploring distributions in 2-dimensional space and in time with separate kernels, and tools for identifying appropriate bandwidths for these.

An implementation of the blocking algorithm KLSH in Steorts, Ventura, Sadinle, Fienberg (2014) <DOI:10.1007/978-3-319-11257-2_20>, which is a k-means variant of locality sensitive hashing. The method is illustrated with examples and a vignette.

This package performs variable selection for many types of L1-regularised regressions using the revisited knockoffs procedure. This procedure uses a matrix of knockoffs of the covariates independent from the response variable Y. The idea is to determine if a covariate belongs to the model depending on whether it enters the model before or after its knockoff. The procedure suits for a wide range of regressions with various types of response variables. Regression models available are exported from the R packages glmnet and ordinalNet'. Based on the paper linked to via the URL below: Gegout A., Gueudin A., Karmann C. (2019) <arXiv:1907.03153>.

Computes Khattree-Bahuguna's univariate and multivariate skewness, principal-component-based Khattree-Bahuguna's multivariate skewness. It also provides several measures of univariate or multivariate skewnesses including, Pearsonâ s coefficient of skewness, Bowleyâ s univariate skewness and Mardia's multivariate skewness. See Khattree, R. and Bahuguna, M. (2019) <doi: 10.1007/s41060-018-0106-1>.