###############################################################################
## NEWS to package MKinfer
###############################################################################

###########################################################
## Version 1.2
###########################################################
- changed print functions for perm.htest and boot.htest
- modified computation of p values for perm.t.test and boot.t.test for extreme
  cases
- removed DESCRIPTION table from package Rd file.
- added functions to compute (multivariate) intersection-union z- and t-test


###########################################################
## Version 1.1
###########################################################
- madCI was modified for method = "boot" and is no longer based on function
  quantileCI, but directly on function mad. It should now give better 
  confidence coverages.
- minor modification in mi.t.test: number of imputations is printed
- added method for S3 class amelia to mi.t.test (from package Amelia)
- added method for S3 class mids to mi.t.test (from package mice)
- added mi.wilcox.test for multiple imputation Wilcoxon tests (based on 
  median p rule)


###########################################################
## Version 1.0
###########################################################
- changed depends from R(>= 3.5.0) to R(>= 4.0.0). As detected by GitHub user
  PaulCristina, functions pairwise.ext.t.test and pairwise.wilcox.exact make
  use of function deparse1, which was added in R 4.0.0.
- now uses deparse1 instead of deparse also in hsu.t.test, perm.t.test, 
  boot.t.test and rm.oneway.test.
- corrected a minor bug regarding data names in rm.oneway.test.
- minor modification of quantileCI for exact confidence intervals and small 
  or high quantiles.
- added parametric and nonparametric Bland-Altman plots including approximative,
  exact and bootstrap confidence intervals.
- added dataset of systolic blood pressure (mm Hg) measurements made using 
  a standard arm cuff and a finger monitor used in Bland and Altman (1995).
- switched back to exhaustive search in the case of quantileCI now using 
  outer instead of a double for loop.


###########################################################
## Version 0.9
###########################################################
- corrected bug in pairwise.wilcox.exact and pairwise.ext.t.test.


###########################################################
## Version 0.8
###########################################################
- simplified computation of quantileCI for method exact according to 
  Section 6.10.2 of Hedderich and Sachs (2020)
- enabled parallel computing for bootstrap confidence intervals
- added pairwise.wilcox.exact and pairwise.ext.t.test


###########################################################
## Version 0.7
###########################################################
- introduced na.rm to volcano
- introduced new parameter corr to imputeSD
- changed http to https
- added some example code for using mice together with mi.t.test to vignette
- added function p2ses for computing a standardized effect size (SES) for 
  permutation tests.


###########################################################
## Version 0.6
###########################################################
- removed an unnecessary print in function binomDiffCI
- use lgamma instead of gamma to compute variance of SD (inside normCI)
- modified imputeSD such that correlations < -1 or > 1, which may occur, are 
  not used for the imputations, but are set to -1 or 1, respectively.


###########################################################
## Version 0.5
###########################################################
- corrected bug in wilson method in binomDiffCI (alpha was not passed through)
  detected by Takuya Yanagida.
- attribute conf.level was not set in meanCI and sdCI.


###########################################################
## Version 0.4
###########################################################
- changed return value of imputeSD
- added volcano plot


###########################################################
## Version 0.3
###########################################################
- corrected minor bug for wald-cc in binomCI and binomDiffCI
- corrected minor bug in case of bootstrap confidence intervals (conf.level 
  was not passed through)
- introduced one-sided confidence intervals in all *CI functions
- added references to description and vignette
- added function rm.oneway.test to compute parametric and non-parametric
  one-way ANOVA with repeated measures
  

###########################################################
## Version 0.2
###########################################################
- Added new function binomDiffCI for computing confidence intervals for
  the difference of two binomial proportions.
- Changed import form MKmisc to MKdescr
- Moved function pairwise.fun from MKmisc/MKdescr to MKinfer
- Added bootstrap intervals to functions binomCI and binomDiffCI
- Added bootstrap intervals to function normCI
- Added functions meanCI and sdCI calling function normCI
- Added bootstrap intervals to function normDiffCI
- Added function meanDiffCI calling function normDiffCI
- Added bootstrap intervals to quantileCI, medianCI and madCI
- Added bootstrap interval to cvCI
- Added function boot.t.test to compute bootstrap t-tests
- Added function perm.t.test to compute permutation t-tests


###########################################################
## Version 0.1
###########################################################
- Start of development
- Copied functions binomCI, cvCI, hsu.t.test, imputeSD,
  madCI, medianCI, mi.t.test, normCI, normDiffCI, 
  quantileCI from package MKmisc