NEWS
metaplus 1.0-5 (2024-07-04)
- changed seq() parameter to along.with from along as could result in errors under certain settings
- updated e-mail address
metaplus 1.0-4 (2022-05-13)
- removed donttest from some examples
metaplus 1.0-3
- fixed crash due to changes in rma procedure requiring missing rather than NULL
- used inherits rather than class for checking class
- updated e-mail address
metaplus 1.0-2 (2021-06-12)
- improved use of integrate() function to avoid problems with t-distribution
- removed option for adaptive gauss-hermite as no longer necessary, and slower
- no longer use multiple starting values for vinv, as not necessary
metaplus 1.0-1 (2021-05-30)
- when performing bootstrap for testing for outliers in mixture model, don't use random starting values, just use suitable values based on the values from the null model
metaplus 1.0-0
- use multiple cores for bootstrap and for multiple starting values for mixture model
- change number of sets of starting values to a fixed number
metaplus 0.7-11 (2018-04-01)
- fixed bug in parsing which caused crash in normal profiling
metaplus 0.7-10 (2018-03-09)
- fixed warning from deprecated use of arrays
metaplus 0.7-9 (2016-11-10)
- changed options for integration of t distribution models so hopefully Gauss-Hermite is never needed
- modified Hessian calculations to avoid possible bug
- added warning message about study second parameter
- updated vignette and added code
metaplus 0.7-8 (2016-09-11)
- updated citation to reflect publication in R Journal
- updated vignette
metaplus 0.7-7 (2016-04-10)
- improved fitting for mixture models
- changed ordering of study labels for outlier probability plot to match ordering of forest plot
- added digits to parameters passed through from plot to summary lines
metaplus 0.7-6 (2016-01-05)
- added profile improvement introduced in 0.7-5 for normal models
- added option to perform integration in the t-distribution models using adaptive Gauss-Hermite quadrature to improve accuracy with studies with small standard errors
- correctly name CDP studies
- documentation enhancements
metaplus 0.7-5 (2015-11-02)
- added data parameter to metaplus to indicate where data is, and changed all examples and documentation to use this
- added a print method for metaplus objects which prints summary
- documentation changes
- renamed outlier.probs to outlierProbs and test.outliers to testOutliers to align with standard R naming conventions
- added further improvements to handing of tau^2 close to zero
- if new maxima is found during profiling then update solution (in practice this usually means that the likelihood is almost flat and very messy)
- add check for multimodal profile likelhood and give warning
- return profile so that this can be further examined
metaplus 0.7-4 (2015-07-03)
metaplus 0.7-3
- fixed occasional problems when tau^2 is zero for all distributions
- if likelihood is mis-shapen and CI plot requested then give a warning rather than fail
- check for inadequate number of studies and stop, or warn if only just adequate
- add cehck for appropriate model when calculating outlier probabilities
- add additional imports in NAMESPACE to pass R CMMD check
metaplus 0.7-2 (2015-06-03)
- fixed bug with fitting t-distribution when tau^2 was zero
- update documentation
metaplus 0.7-1 (2015-01-13)
- fixed bug related to singularity of Hessian
- when better solution found when profiling then use this fitted model
metaplus 0.7-0 (2015-01-11)
- improved convergence by using Nelder-Mead if nlminb failed
- used modified bbmle to fix bugs - changes will eventually be incorporated into bbmle
metaplus 0.6-1 (2014-10-14)
- improve profiling of t-distribution models by using multiple starting values
- fix problem when vinv=0 correctly
- fix documentation errors
- include examples vignette
metaplus 0.6-0 (2014-08-20)