Iterative Solvers for (Sparse) Linear System of Equations (Rlinsolve)
Solving a system of linear equations is one of the most fundamental computational problems for many fields of mathematical studies, such as regression problems from statistics or numerical partial differential equations. We provide basic stationary iterative solvers such as Jacobi, Gauss-Seidel, Successive Over-Relaxation and SSOR methods. Nonstationary – or, Krylov subspace methods are also provided; Conjugate Gradient, Conjugate Gradient Squared, Biconjugate Gradient, and Biconjugate Gradient Stabilized methods. Sparse matrix computation is also supported in that solving large and sparse linear systems can be manageable using ‘Matrix’ package along with ‘RcppArmadillo’. For a more detailed description, see a book by Saad (2003) <doi:10.1137/1.9780898718003>.

sf’-Compatible Interface to ‘Google Maps’ APIs (mapsapi)
Interface to the ‘Google Maps’ APIs: (1) routing directions based on the ‘Directions API’, returned as ‘sf’ objects, either as single feature per alternative route, or a single feature per segment per alternative route, (2) travel distance or time matrices based on the ‘Distance Matrix API’.

Multiple Curve Comparisons Using Parametric Bootstrap (curvecomp)
Performs multiple comparison procedures on curve observations among different treatment groups. The methods are applicable in a variety of situations (such as independent groups with equal or unequal sample sizes, or repeated measures) by using parametric bootstrap. References to these procedures can be found at Konietschke, Gel, and Brunner (2014) <doi:10.1090/conm/622/12431> and Westfall (2011) <doi:10.1080/10543406.2011.607751>.

A Hybrid Filter-Wrapper Feature Selection Method (HybridFS)
A hybrid method of feature selection which combines both filter and wrapper methods. The first level involves feature reduction based on some of the important Filter methods while the second level involves feature subset selection as in a wrapper method. Experimental results show that this hybrid feature selection algorithm simplifies the feature selection process effectively and obtains higher classification accuracy, reduced processing time and improved data handling capacity than other feature selection algorithms.

Estimate the Population Size for the Mb Capture-Recapture Model (OBMbpkg)
Applies an objective Bayesian method to the Mb capture-recapture model to estimate the population size N. The Mb model is a class of capture-recapture methods used to account for variations in capture probability due to animal behavior. Under the Mb formulation, the initial capture of an animal may effect the probability of subsequent captures due to their becoming ‘trap happy’ or ‘trap shy.’

Test for and Identify Categorical or Continuous Values (catcont)
Methods and utilities for classifying vectors as categorical or continuous.