Changes in version 1.4-0 (2026-05-29) Features o Add invchol.syMatrices, cov2invchol, and cov2chol. o Avoid swapping data (row <-> column-wise storage) in a number of places. Bugfixes o Cleaned up the class structure and corrected constructors for syMatrices. Other o After more than 15 years on R-forge, mvtnorm is now hosted at codeberg. Thank you, WU Vienna, for the long-term support! Changes in version 1.3-8 Features o More careful checks of inputs. Bugfixes o logLik wasn't able to handle missing, continuous, and interval-censored observations at once. o lpmvnorm failed when a non-missing invcholmean wasn't a matrix with the same dimensions as lower and upper. Changes in version 1.3-7 (2026-04-15) Features o Minor performance improvements in sldmvnorm. o Speed-up switching from row- to column major order and vice versa. o Add invcholmean argument where possible. o Improve documentation regarding marginal and conditional distributions. o Ship nuweb source file in inst/nuweb Bugfixes o simulate ignored its seed argument. Changes in version 1.3-6 (2026-03-15) Bugfixes o sldmvnorm returned an incorrect log-likelihood in the presence of means. o Observations with very small log-probabilities are now treated the same in lpmvnorm and slpmvnorm. o This reduced differences observed on machines w/o long doubles considerably; still there were minor diffs in the package vignette which we now prevent from showing up. Changes in version 1.3-5 (2026-03-11) Documentation o Use REFERENCES.bib. Changes in version 1.3-4 Features o mvnorm gains a invcholmean argument allowing a parameterisation leading to convex optimisation problems. o Added zero.print = "." argument for printing lower triangular matrices. Bugfixes o condDist() sometimes failed to return a scale slot. Documentation o Explain how to use condDist() to obtain regression coefficients for an example involving means. Changes in version 1.3-3 (2025-01-10) Bugfixes o Miwa(steps = 4097) are allowed in R, the C code assumed steps <= 4096, now steps == 4097 is possible. Triggered by BDR's gcc-UBSAN check on package OptimalGoldstandardDesign. Changes in version 1.3-2 (2024-11-04) Features o Explain and illustrate that omitting data dimensions produces marginal log-likelihoods in Chapter 7. o Add low-level functionality for reduced rank covariances in new Chapter 8. Bugfixes o invcholD and Dchol avoided very small diagonals; now they also avoid very large diagonals (such that inverting the result doesn't give too small diagonals). o Fix partial argument matches, reported by KH. Changes in version 1.3-1 (2024-09-03) Bugfixes o Challenging numeric test of score function at ML estimate failed on Mac M1. Changes in version 1.2-6 (2024-08-17) Features o Add logdet() function for computing log-determinants from lower triangular matrices. o Speed up solve(, ...). Bugfixes o Avoid unnecessary copying of large data objects. o -1/2 -> -1 for det(C) in lmvnorm_src vignette; spotted by Kurt Hornik o Use M\_PI. Changes in version 1.2-5 (2024-05-21) Features o Update the Using mvtnorm package vignette and references therein. o Speed up Mult(, transpose = TRUE). o Speed up ldmvnorm(). o Add constructor syMatrices() for multiple symmetric matrices and as.syMatrices() for coercion from ltMatrices. Bugfixes o Fix segmentation fault or unnecessary error or warning + approximation for algorithm = Miwa() or TVPACK(), in case dimension reduction to one-dimensional is possible, e.g., for pmvnorm(lower = rep(-Inf,3), upper = c(-1, Inf, Inf), sigma = diag(3), algorithm = Miwa()) # or pmvnorm(lower = c(-Inf,-Inf), upper = c(- 1, Inf), sigma=diag(2), algorithm = TVPACK()) Changes in version 1.2-4 (2023-11-27) Bugfixes o Remove empty print statement from miwa.c Changes in version 1.2-3 (2023-08-25) Features o Allow to change rnorm in rmvnorm, feature request by Ralf Stubner. Bugfixes o Fix variable declarations in tvpack.f as reported by Intel compilers icx/ipcx/ifx from oneAPI 2023.2.0 and oneMKL 2023.2.0, thanks to BDR Changes in version 1.2-2 (2023-06-08) Bugfixes o Fix overflow problem reported by ASAN. Documentation o Be even more careful inverting / computing Cholesky factors for hessians (M1 and macos-arm64). Changes in version 1.2-1 (2023-06-04) Bugfixes o Avoid attempts to allocate zero length arrays for one-dimensional problems. Documentation o Catch M1mac problems with inverting hessians. Changes in version 1.2-0 (2023-06-02) o New lpmvnorm() function for computing multivariate normal log-likelihoods for interval-censored observations. o New slpmvnorm() function for computing the corresponding score function for interval-censored observations. o New ldmvnorm() function for computing multivariate normal log-likelihoods for exact observations. o New sldmvnorm() function for computing the corresponding score function for exact observations. o New ldpmvnorm() function for computing multivariate normal log-likelihoods for a mix of exact and interval-censored observations. o New sldpmvnorm() function for computing the corresponding score function for a mix of exact and interval-censored observations. o New class ltMatrices representing multiple lower triangular matrices, with some useful methods. o New package vignette lmvnorm_src describing these new features. o pmvnorm(), qmvnorm(), pmvt(), and qmvt() gain a seed argument. o Regression tests were improved. o Internal standardization is described better in the docs, suggested by Chris Wymant. o New NEWS file, the old file containing information up to version 1.1-3 is available as NEWS.old.