- BUG-FIX -
perc_meandiff
calculated the degree of freedom incorrectly for equal variance differences (see #106) - addition of Cronbach's alpha to the shiny app
- addition of Cronbach's alpha method
- notes in helpfile for
prec_mean
andprec_meandiff
that the former is suitable for paired mean differences. Comparable changes in the shiny app.
- extra note for
prec_sens
andprec_spec
regarding then
in the output - minor clarification for
prec_icc
regarding assumption of no interaction between rater and subject (see #97) - address warnings with prec_prop (see #94)
- Update URL
- minor changes requested by CRAN to update to HTML5 helpfiles.
- addition of vignette
- BREAKING CHANGE - prec_mean argument changed from 'mu' to 'mean'.
- logo
- installation via universe
- addition of citation info
- update SCTO figures
- minor formatting of URLs and references to pass CRAN checks
- bug fix in shiny app for sens and spec
- more examples, better described
- add extra input validation to all functions
- add reset buttons to shiny app
- correct copy/paste error on ICC page
- various changes to ensure that all functions support vectors for different scenarios
- minor changes to readme and shinyapp (corrections of references/typos)
- more minor changes for CRAN
- minor changes requested by CRAN
- first version for CRAN
- minor clarifications to options/descriptions
- minor changes to shiny app
- initial 'final' version
- addition of wrappers for
prec_lr
(prec_pos_lr
,prec_neg_lr
) to simplify positive/negative LRs
- addition of method for likelihood ratios
prec_lr
-
addition of function for Cohen's kappa
-
update shiny app to include kappa
-
POSSIBLE BREAKING CHANGE: arguments in
prec_rateratio
renamed from*_exp
and*_control
to*1
and*2
for consistency with other functions
- addition of shiny app and pkgdown
-
prec_sens
andprec_spec
allow prev and conf.width -
multiple notes allowed in print method
-
add contributing guidelines
-
prec_sens_spec
removed. Confidence intervals and sample sizes are quite different to other methods. A two step approach usingprec_sens
andprec_spec
is instead recommended
-
addition of various tests
-
addition of rate ratio method