argument int removed from plotfit(): can no longer launch shiny apps from the plotfit() command for plotting distributions. Use elicit() and elicitMultiple() instead for interactive plotting.
elicitation report files (R Markdown documents) now include plots of fitted distributions.
Bugs fixed in elicitDirichlet() - code wouldn’t run with more than three categories.
fitdist() has a new argument for excluding log t and mirror log t when identifying best fit (default is FALSE).
Plots can be downloaded from shiny apps as .png files.
New (negatively skewed) distributions: mirror gamma, mirror lognormal, and mirror log t. These all fit distributions to (upper - X).
Better handling of expert names: elicitMultiple(): can now click on and edit expert names; names are displayed in all plots. plotfit() will now use the expert names, if provided in fitdist(). Can also specify names in plotTertiles() and plotQuartiles().
elicitMultiple(): can now control axes limits in quartile and tertile plot.
new function rlinearpool() for sampling from a weighted linear pool
bug fixed in copulaSample. Was rounding samples to 3 s.f. Increased to 8. Mistake in help file corrected, regarding syntax for distribution names.
elicitExtension(): can now upload a sample from the distribution of the extension variable, instead of eliciting a distribution.
plotfit(): additional argument returnPlot (default is FALSE) will also return the plot as a ggplot object.
new app for the extension method: elicitMixture(), for discrete extension variables.
bug fixed in elicitExtension(): when plotting conditional densities for the logit link function, x-axis limits now restricted to 0 and 1.
bug fixed in plotfit(): will now correctly display a single distribution for a selected expert when requested, if multiple distributions have been elicited.
plinearpool() and qlinearpool(): can now directly specify different distribution types for each expert to use in the linear pool.
fitdist(): extra argument expertnames, for specifying row names in the various outputs.
elicit() app: can now report fitted probabilities as well as fitted quantiles, and can change x-axis label
bug fixed: switched from class(x) == “foo” to inherits(x, “foo”), to avoid assumption length(class(x)) == 1
new app for the extension method: elicitExtension(). New command line functions for the extension method are plotConditionalDensities(), plotConditionalMedianFunction() and sampleMarginalFit().
makeCDFPlot() function is now exported: plots the elicited cumulative probabilities, and fitted cumulative distribution functions.
elicitMultiple() app: can now enter judgements with the roulette method, and save/load judgements as .csv files
column names changed in output of feedback(), fitdist() and sampleFit() to be consistent: “normal”, “t”, “gamma”, “lognormal”, “logt”, “beta”, “hist”
roulette() has been removed, and the roulette method is now available within elicit()
Extra argument percentages in plotfit() and plotTertiles() for using percentage scale on x-axis
New function sampleFit(), for generating samples from fitted distributions.
Minor change to fitDirichlet(), to allow marginal elicitation fits to be specified as a single list.
Update to fitprecision(): interval used in the proportion method can now be a tail area of the population distribution
New shiny app elicitBivariate() for eliciting bivariate distributions using a Gaussian copula
Significant update to elicit() shiny app: can now switch between multiple methods within the same app
New shiny app elicitMultiple() for fitting individual distributions to multiple experts’ judgements
Bugs fixed: plinearpool() now chooses the best fitting distribution for each expert if argument d = “best” is specified. Correctly handles probabilities for log-t, where x is below lower limit.
Bugs fixed: qlinearpool() could return NA in some cases if argument d = “best” was specified: now fixed. Correctly handles probabilities for log-t, where x is below lower limit. Minor improvement to accuracy in estimated quantiles: finer grid used in linear interpolation of the quantile function.
New function: generateReport(): renders an Rmarkdown document to give formulae and parameter values for all the fitted distributions
New function: condDirichlet(), for viewing conditional distributions from elicited Dirichlet distributions
New functions: plotQuartiles() and plotTertiles(), for displaying individuals quartiles/tertiles elicited from a group of experts
New functions: elicitQuartiles() and elicitTertiles(): shiny apps for eliciting with the quartile and tertile methods
elicit() and roulette() functions now both return the elicited values and results as objects of class “elicitation”
Bug fixed: ensure solid line used for linear pool when plotting. Option in plotfit added to plot all individual densities with same colour, to simplify legend.
New function: linearPoolDensity, for extracting density values from the linear pool.
Bug fixed: can now accept more than 26 experts.
Bug fixed: qlinearpool/plinearpool now works with log t distributions.
New function: elicitHeterogen, for eliciting prior for variance of random effects in meta-analysis
Bug fixed: can fit (and plot) distributions bounded below when lower limit is negative
Bug fixed: roulette method shiny interface works with non-integer bin boundaries
Accept non-decreasing probabilities in elicited judgements, rather than only strictly increasing probabilities
Can specify own axes labels in the plotfit command with arguments xlab and ylab
Update to Multivariate-normal-copula.Rmd vignette, to match update to GGally
Bug fixed: interactive plots now work for plotting individual distributions for multiple experts
Bug fixed: plotting best fitting individual distributions for multiple experts
Roulette elicitation method now implemented using shiny
New functions fitDirchlet and feedbackDirichlet for eliciting Dirichlet distributions
New functions copulaSample and elicitConcProb for eliciting dependent distributions using multivariate normal copulas
New function compareIntervals for comparing fitted intervals for individual distributions from multiple experts
Change to expert.names from numbers to letters in fitdist
Vignettes added: overview of SHELF, eliciting a Dirichlet distribution, eliciting a bivariate distribution with a bivariate normal copula
Change in fitdist to starting values in optimisation: will now check for exact fits if only two probabilities elicited
New functions added for eliciting beliefs about uncertain population distributions: cdffeedback, cdfplot, fitprecision, pdfplots