Estimation and Hypothesis Testing for Threshold Regression (chngpt)
Threshold regression models are also called two-phase regression, broken-stick regression, split-point regression, structural change models, and regression kink models. Methods for both continuous and discontinuous threshold models are included, but the support for the former is much greater. This package is described in Fong, Huang, Gilbert and Permar (2017) chngpt: threshold regression model estimation and inference, BMC Bioinformatics, in press, <DOI:10.1186/s12859-017-1863-x>.

Interface to ‘H2O4GPU’ (h2o4gpu)
Interface to ‘H2O4GPU’ <https://…/h2o4gpu>, a collection of ‘GPU’ solvers for machine learning algorithms.

Landscape Utility Toolbox (landscapetools)
Provides utility functions to complete tasks involved in most landscape analysis. It includes functions to coerce raster data to the common tibble format and vice versa, it helps with flexible reclassification tasks of raster data and it provides a function to merge multiple raster. Furthermore, ‘landscapetools’ helps landscape scientists to visualize their data by providing optional themes and utility functions to plot single landscapes, rasterstacks, -bricks and lists of raster.

Bindings for ‘Mapbox’ Ear Cutting Triangulation Library (decido)
Provides constrained triangulation of polygons. Ear cutting (or ear clipping) applies constrained triangulation by successively ‘cutting’ triangles from a polygon defined by path/s. Holes are supported by introducing a bridge segment between polygon paths. This package wraps the ‘header-only’ library ‘earcut.hpp’ <https://…/earcut.hpp.git> which includes a reference to the method used by Held, M. (2001) <doi:10.1007/s00453-001-0028-4>.

Variable Selection for Missing Data (TVsMiss)
Use a regularization likelihood method to achieve variable selection purpose, can be used with penalty lasso, smoothly clipped absolute deviations (SCAD) and minimax concave penalty (MCP). Tuning parameter selection techniques include cross validation (CV), Bayesian information criterion (BIC) (low and high), stability of variable selection (sVS), stability of BIC (sBIC), and stability of estimation (sEST). More details see Zhao, Jiwei, Yang Yang, and Yang Ning (2018) <arXiv:1703.06379> ‘Penalized pairwise pseudo likelihood for variable selection with nonignorable missing data.’ Statistica Sinica.