Changes in lqmm

=============================================

Version 1.0 (April 29, 2012)

- Package lqmm on CRAN

=============================================

Version 1.01 (May 30, 2012)

- Added functions
- Fixed bugs in ll_h_R and ll_h_d (lqmm.c)
- Amended cov.sel (lqmm.R) and psi_mat (lqmm.c)

=============================================

Version 1.02 (September 15, 2012)

- Changed lqmControl and lqmmControl parameters (less stringent convergence thresholds) (lqmm.R)
- Changed quadrature default in lqmm (lqmm.R)
- Changed to the function boot.lqmm (lqmm.R) which provides bootstrapped standard errors of the coefficients estimates. The starting values used in each bootstrap sample are by default those from a least squares fit (startQR = FALSE). In previous versions of the package, starting values were taken from the fitted object (startQR = TRUE) to speed up convergence in each bootstrap sample. However, this causes the standard errors to be under-estimated

=============================================

Version 1.03 (April 30, 2013)

- Improved documentation
- Changed default lqmmControl(method = "df") to lqmmControl(method = "gs") (lqmm.R)
- References updated
- Fixed bug in covHandling (lqmm.R)
- Produce warning when likelihood ratio test is negative

=============================================

Version 1.04 (November 1, 2013)

- Changed license from GPL >= 3 to GPL >= 2
- Added vignette lqmm_manual.pdf
- Improved code and documentation
- Corrected bug in lqm.counts (in previous package versions standard errors were NOT divided by sqrt(n))
- New functions for lqm.counts objects

=============================================

Version 1.5 (May 1, 2014)

- Changed versioning
- S3 methods registered
- Introduced new generic functions boot, extractBoot
- Function raneff.lqmm renamed ranef.lqmm
- Function cov.lqmm renamed VarCorr.lqmm
- Imported nlme generic ranef and VarCorr

=============================================

Version 1.5.1 (December 10, 2014)

- Changed maintainer's email address
- Corrected bug in qal
- Added constraint on tau in asymmetric Laplace distribution functions
- Added warning when user defines sigma < 0 in asymmetric Laplace distribution functions


=============================================

Version 1.5.2 (September 21, 2015)

- Corrected bug in predict.lqm
- Corrected bug in varAL
- Removed rounding of values of 'tau' from lqm and lqmm
- New lqm routine ("gs2") more stable with larger datasets. Default is "gs1". See ?lqmControl for details
- New function predint.lqmm to calculate prediction intervals in lqmm
- Replaced ChangeLog with NEWS (which is more appropriate)
- Acknowledgements: thanks to Nathan Pace for reporting bugs and other issues

=============================================

Version 1.5.3 (January 28, 2016)

- Updated function VarCorr for compatibility with nlme 3.1-123

=============================================

Version 1.5.4 (April 7, 2018)

- Fixed bug in createLaguerre
- lqmgsR: changed gradient (no division by n)
- Added smooth approximation loss function
- Registered dll routines
- Fixed CRAN message 'Note: break used in wrong context: no loop is visible'
- Removed LRT from lqmm summary

=============================================

Version 1.5.5 (December 12, 2019)

- Fixed CRAN error related to _R_CLASS_MATRIX_ARRAY_ starting from R 4.0.0
- Fixed bug in boot.lqm
- Fixed bug in boot.lqmm

=============================================

Version 1.5.6 (November 4, 2021)

- CRAN request to define USE_FC_LEN_T in headers, obligatory starting from R 4.2.0
- Added newdata argument in predict.lqmm
- Changed maintainer's email

=============================================

Version 1.5.7 (November 25, 2021)

- fixed bug in boot.lqmm (try(..., silent = TRUE))
- Fixed bug in predict.lqmm (thanks to Max Gordon)

=============================================

Version 1.5.8 (April 5, 2022)

- CRAN request to fix BLAS