NEWS

---

crch 1.2-1 (2024-09-16)

• Due to a bug in the crps() computation for censored and truncated Student's T distributions in the scoringRules package, some example output in ?CensoredStudentsT and ?TruncatedStudentsT was in error. The problem has been fixed in version 1.1.3 of scoringRules and reference output in crch was updated.

crch 1.2-0 (2024-08-22)

• Improved implementations of expectation and variance of censored and truncated normal distributions (contributed by Ioannis Kosmidis).

• Turned vignette("crch", package = "crch") from Sweave into Quarto vignettes. Some improvements/updates in the text.

• New package web page (via altdoc/quarto) at https://topmodels.R-Forge.R-project.org/crch/

• Achim Zeileis takes over maintenance from Jakob W. Messner. Added Ioannis Kosmidis and Georg J. Mayr as contributors.

crch 1.1-2 (2023-03-27)

• Fixed name of first argument in crps() method to be consistent with the generic.

crch 1.1-1 (2022-09-09)

• Added methods for is_discrete() and is_continuous() for the new distributions3 objects.

• Replaced deprecated C function finite() with isfinite() (again).

crch 1.1-0 (2022-09-07)

• Added support for distributions3 workflows for censored and truncated normal, logistic, and Student's t distributions: CensoredNormal(), TruncatedNormal(), CensoredLogistic(), TruncatedLogistic(), CensoredStudentsT(), TruncatedStudentsT(). See the corresponding manual pages for examples illustrating how to work with the distributions in practice, for computing moments, probabilities, densities, simulating random values, etc.

• Added prodist() method for extracting the distributions3 objects above from fitted crch objects, either in-sample or out-of-sample.

• Bug fix in the computation of the mean of censored or truncated logistic distributions with large (or infinite) censoring/truncation points.

crch 1.0-4 (2019-09-03)

• Replaced deprecated C function finite() with isfinite().

crch 1.0-3 (2019-05-08)

• Updated contact information.

crch 1.0-1 (2018-01-23)

• Added argument type to crch() which can be set to "crps" for parameter estimation with minimum CRPS instead of maximum likelihood.

• Added S3 method for crps() from scoringRules for crch objects.

• Improvements for the predict() method:

  • New types "parameter", "density", "probability", and "crps".
  • With type = "response" now the expected value and not the location parameter is returned (not equal for censored and truncated distributions). For better backward compatibility, the default type is set to "location".

• Added pit(), rootogram(), and simulate() methods for crch objects.

• Changed argument names mean and sd to location and scale in logistic and Student's t distribution functions

• Added new function crch.stabsel() for stability selection based on
  crch.boost.fit(). Some S3 methods for the returned class stabsel.crch are also provided.

crch 1.0-0 (2016-10-18)

• New release accompanying the R Journal paper: Heteroscedastic Censored and Truncated Regression with crch by Messner, Mayr, and Zeileis. See also citation("crch").

• Added estfun() method for crch objects

crch 0.9-2 (2016-02-29)

• The crch() function now supports coefficient optimization by boosting to automatically select the most relevant input variables in high-dimensional data settings. Extractor and plotting functions for corresponding crch.boost objects are also available.

• Transferred functions to estimate density, distribution, score, and Hessian matrices to C code to accelerate coefficient optimization.

• Added option to crch() to avoid computation of covariance matrix.

• Added left and right arguments to predict.crch() and predict.crch.boost() to allow quantile predictions for non-constant censoring or truncation points.

crch 0.9-1 (2015-08-06)

• Added model.matrix() and model.frame() methods for crch objects

• Bug fix in predict.crch(): In previous versions predictions for models with other link functions than the log gave wrong results

crch 0.9-0 (2015-06-27)

• Added vignette to introduce the crch() function with some theoretical background and an illustrating example: vignette("crch", package = "crch")

• The crch() function now also supports truncated responses. Furthermore added a wrapper function trch() to fit truncated regression models.

• crch(): Analytical gradients and Hessian matrices are provided for most models to speed up maximum likelihood optimization (not available for Student's t distribution with degrees of freedom estimation).

• crch(): For the scale model a link function can now be specified ("log", "identity", or "quadratic"). In previous version only the log was supported.

• Added functions for probability density, cumulative distribution, random numbers, and quantiles for censored and truncated normal, logistic, and Student's t distributions.

• The residuals() method for crch objects now also provides quantile residuals (Dunn and Smyth 1996).

• Added update() method for crch objects.

crch 0.1-0 (2013-10-22)

• First official release of the package on CRAN. See citation("crch") for the accompanying manuscripts. Note that the interface of both crch() and hxlr() is still under development and might change in future versions of the package.