NEWS.md
makeDoe()
function, including an improved print out of the summary of the design (D-efficiency & balanced or not), as well as optional arguments to conduct a search of designs.evaluateDoe()
function which returns a list of information about a given design, including the D-efficiency and whether or not the design is balanced.recodeDesign()
to recodeDoe()
and made it more robust to different data types in levels
.parallel::mcapply()
in estimateModels()
that was introduced in 0.0.7 with just lapply()
as the parallel
approach is not supported in Windows. Will add more robust parallel implementation in next package update.Bug fixes:
simulateChoices()
with a provided utility model, the function was incorrectly categorizing categorical variables with only 2 levels as being numeric. This is now fixed.predict.logitr()
method introduced in {logitr} v0.4.0.estimateModels()
function.print.cjtools()
method for a prettier printing of the models object that is returned from the estimateModels()
function.Improved how the models are estimated in estimateModels()
for increased speed using a nested approach with {data.table}.
Updated functions to be consistent with some breaking changes introduced when {logitr} was updated from 0.2.0 to 0.3.0, which is necessary since {logitr} is the underlying package used to estimate models in the power analyses.
estimateModels()
and getModelResults()
functions for conducting a power analysis.simulateChoices()
function.makeSurvey()
function.makeDoe()
and makeSurvey()
functions to enable grouped designs.recodeDesign()
function.Interactions!