This package provides functions to upload vectorial data and derive landscape connectivity metrics in habitat or matrix systems. Additionally, includes an approach to assess individual patch contribution to the overall landscape connectivity, enabling the prioritization of habitat patches. The computation of landscape connectivity and patch importance are very useful in Landscape Ecology research. The metrics available are: number of components, number of links, size of the largest component, mean size of components, class coincidence probability, landscape coincidence probability, characteristic path length, expected cluster size, area-weighted flux and integral index of connectivity. Pascual-Hortal, L., and Saura, S. (2006) <doi:10.1007/s10980-006-0013-z> Urban, D., and Keitt, T. (2001) <doi:10.2307/2679983> Laita, A., Kotiaho, J., Monkkonen, M. (2011) <doi:10.1007/s10980-011-9620-4>.

"Learning with Subset Stacking" is a supervised learning algorithm that is based on training many local estimators on subsets of a given dataset, and then passing their predictions to a global estimator. You can find the details about LESS in our manuscript at <arXiv:2112.06251>.

This package provides a word embeddings-based semi-supervised model for document scaling Watanabe (2020) <doi:10.1080/19312458.2020.1832976>. LSS allows users to analyze large and complex corpora on arbitrary dimensions with seed words exploiting efficiency of word embeddings (SVD, Glove). It can generate word vectors on a users-provided corpus or incorporate a pre-trained word vectors.

Robust test(s) for model diagnostics in regression. The current version contains a robust test for functional specification (linearity). The test is based on the robust bounded-influence test by Heritier and Ronchetti (1994) <doi:10.1080/01621459.1994.10476822>.

When the values of the outcome variable Y are either 0 or 1, the function lsm() calculates the estimation of the log likelihood in the saturated model. This model is characterized by Llinas (2006, ISSN:2389-8976) in section 2.3 through the assumptions 1 and 2. The function LogLik() works (almost perfectly) when the number of independent variables K is high, but for small K it calculates wrong values in some cases. For this reason, when Y is dichotomous and the data are grouped in J populations, it is recommended to use the function lsm() because it works very well for all K.

This computes Lipinski Rule of Five parameters and offers visualization for drug discovery. It analyzes molecular properties like molecular weight, hydrogen bond donors, acceptors, and ALogP, providing histograms and pass/fail status plots for efficient compound evaluation, aiding in drug development.

This package provides a classification tree method that uses Uncorrelated Linear Discriminant Analysis (ULDA) for variable selection, split determination, and model fitting in terminal nodes. It automatically handles missing values and offers visualization tools. For more details, see Wang (2024) <doi:10.48550/arXiv.2410.23147>.

Fast estimation of multinomial (MNL) and mixed logit (MXL) models in R. Models can be estimated using "Preference" space or "Willingness-to-pay" (WTP) space utility parameterizations. Weighted models can also be estimated. An option is available to run a parallelized multistart optimization loop with random starting points in each iteration, which is useful for non-convex problems like MXL models or models with WTP space utility parameterizations. The main optimization loop uses the nloptr package to minimize the negative log-likelihood function. Additional functions are available for computing and comparing WTP from both preference space and WTP space models and for predicting expected choices and choice probabilities for sets of alternatives based on an estimated model. Mixed logit models can include uncorrelated or correlated heterogeneity covariances and are estimated using maximum simulated likelihood based on the algorithms in Train (2009) <doi:10.1017/CBO9780511805271>. More details can be found in Helveston (2023) <doi:10.18637/jss.v105.i10>.

Local Polynomial Regression with Ridging.

Fits and tests logistic joinpoint models.

Convenient aliases for common ways of misspelling the base R function length(). These include every permutation of the final three letters.

Error in a binary dependent variable, also known as misclassification, has not drawn much attention in psychology. Ignoring misclassification in logistic regression can result in misleading parameter estimates and statistical inference. This package conducts logistic regression analysis with misspecification in outcome variables.

An HTML widget that randomly tours 2D projections of numerical data. A random walk through projections of the data is shown. The user can manipulate the plot to use specified axes, or turn on Guided Tour mode to find an informative projection of the data. Groups within the data can be hidden or shown, as can particular axes. Points can be brushed, and the selection can be linked to other widgets using crosstalk. The underlying method to produce the random walk and projection pursuit uses Langevin dynamics. The widget can be used from within R, or included in a self-contained R Markdown or Quarto document or presentation, or used in a Shiny app.

Lake morphometry metrics are used by limnologists to understand, among other things, the ecological processes in a lake. Traditionally, these metrics are calculated by hand, with planimeters, and increasingly with commercial GIS products. All of these methods work; however, they are either outdated, difficult to reproduce, or require expensive licenses to use. The lakemorpho package provides the tools to calculate a typical suite of these metrics from an input elevation model and lake polygon. The metrics currently supported are: fetch, major axis, minor axis, major/minor axis ratio, maximum length, maximum width, mean width, maximum depth, mean depth, shoreline development, shoreline length, surface area, and volume.

Probabilistic record linkage without direct identifiers using only diagnosis codes. Method is detailed in: Hejblum, Weber, Liao, Palmer, Churchill, Szolovits, Murphy, Kohane & Cai (2019) <doi: 10.1038/sdata.2018.298> ; Zhang, Hejblum, Weber, Palmer, Churchill, Szolovits, Murphy, Liao, Kohane & Cai (2021) <doi: 10.1093/jamia/ocab187>.

Latent Class Analysis of phenotypic measurements in pedigrees and model selection based on one of two methods: likelihood-based cross-validation and Bayesian Information Criterion. Computation of individual and triplet child-parents weights in a pedigree is performed using an upward-downward algorithm. The model takes into account the familial dependence defined by the pedigree structure by considering that a class of a child depends on his parents classes via triplet-transition probabilities of the classes. The package handles the case where measurements are available on all subjects and the case where measurements are available only on symptomatic (i.e. affected) subjects. Distributions for discrete (or ordinal) and continuous data are currently implemented. The package can deal with missing data.

Set of tools for analyzing vertical fuel continuity at the tree level using Airborne Laser Scanning data. The workflow consisted of: 1) calculating the vertical height profiles of each segmented tree; 2) identifying gaps and fuel layers; 3) estimating the distance between fuel layers; and 4) retrieving the fuel layers base height and depth. Additionally, other functions recalculate previous metrics after considering distances greater than certain threshold. Moreover, the package calculates: i) the percentage of Leaf Area Density comprised in each fuel layer, ii) remove fuel layers with Leaf Area Density (LAD) percentage less than 10, and iii) recalculate the distances among the reminder ones. On the other hand, it identifies the crown base height (CBH) based on different criteria: the fuel layer with the highest LAD percentage and the fuel layers located at the largest- and at the last-distance. When there is only one fuel layer, it also identifies the CBH performing a segmented linear regression (breaking points) on the cumulative sum of LAD as a function of height. Finally, a collection of plotting functions is developed to represent: i) the initial gaps and fuel layers; ii) the fuels base height, depths and gaps with distances greater than certain threshold and, iii) the CBH based on different criteria. The methods implemented in this package are original and have not been published elsewhere.

Locally sparse estimator of generalized varying coefficient model for asynchronous longitudinal data by kernel-weighted estimating equation.

Statistical tests widely utilized in biostatistics, public policy, and law. Along with the well-known tests for equality of means and variances, randomness, and measures of relative variability, the package contains new robust tests of symmetry, omnibus and directional tests of normality, and their graphical counterparts such as robust QQ plot, robust trend tests for variances, etc. All implemented tests and methods are illustrated by simulations and real-life examples from legal statistics, economics, and biostatistics.

This package provides classes and methods that allow the user to manage life table, actuarial tables (also multiple decrements tables). Moreover, functions to easily perform demographic, financial and actuarial mathematics on life contingencies insurances calculations are contained therein. See Spedicato (2013) <doi:10.18637/jss.v055.i10>.

An implementation of logistic normal multinomial (LNM) clustering. It is an extension of LNM mixture model proposed by Fang and Subedi (2020) <arXiv:2011.06682>, and is designed for clustering compositional data. The package includes 3 extended models: LNM Factor Analyzer (LNM-FA), LNM Bicluster Mixture Model (LNM-BMM) and Penalized LNM Factor Analyzer (LNM-FA). There are several advantages of LNM models: 1. LNM provides more flexible covariance structure; 2. Factor analyzer can reduce the number of parameters to estimate; 3. Bicluster can simultaneously cluster subjects and taxa, and provides significant biological insights; 4. Penalty term allows sparse estimation in the covariance matrix. Details for model assumptions and interpretation can be found in papers: Tu and Subedi (2021) <arXiv:2101.01871> and Tu and Subedi (2022) <doi:10.1002/sam.11555>.

This package provides a class that links matrix-like objects (nodes) by rows or by columns while behaving similarly to a base R matrix. Very large matrices are supported if the nodes are file-backed matrices.

Fast implementations to compute the genetic covariance matrix, the Jaccard similarity matrix, the s-matrix (the weighted Jaccard similarity matrix), and the (classic or robust) genomic relationship matrix of a (dense or sparse) input matrix (see Hahn, Lutz, Hecker, Prokopenko, Cho, Silverman, Weiss, and Lange (2020) <doi:10.1002/gepi.22356>). Full support for sparse matrices from the R-package Matrix'. Additionally, an implementation of the power method (von Mises iteration) to compute the largest eigenvector of a matrix is included, a function to perform an automated full run of global and local correlations in population stratification data, a function to compute sliding windows, and a function to invert minor alleles and to select those variants/loci exceeding a minimal cutoff value. New functionality in locStra allows one to extract the k leading eigenvectors of the genetic covariance matrix, Jaccard similarity matrix, s-matrix, and genomic relationship matrix via fast PCA without actually computing the similarity matrices. The fast PCA to compute the k leading eigenvectors can now also be run directly from bed'+'bim'+'fam files.

This package performs approximate GP regression for large computer experiments and spatial datasets. The approximation is based on finding small local designs for prediction (independently) at particular inputs. OpenMP and SNOW parallelization are supported for prediction over a vast out-of-sample testing set; GPU acceleration is also supported for an important subroutine. OpenMP and GPU features may require special compilation. An interface to lower-level (full) GP inference and prediction is provided. Wrapper routines for blackbox optimization under mixed equality and inequality constraints via an augmented Lagrangian scheme, and for large scale computer model calibration, are also provided. For details and tutorial, see Gramacy (2016 <doi:10.18637/jss.v072.i01>.