Implements various methods for eliciting a probability distribution for a single parameter from an expert or a group of experts. The expert provides a small number of probability judgements, corresponding to points on his or her cumulative distribution function. A range of parametric distributions can then be fitted and displayed, with feedback provided in the form of fitted probabilities and percentiles. A graphical interface for the roulette elicitation method is also provided. For multiple experts, a weighted linear pool can be calculated. Also includes functions for eliciting beliefs about population distributions.
| Version: | 1.2.1 |
| Depends: | R (≥ 3.3.1) |
| Imports: | ggplot2, grid, shiny, stats, graphics, tidyr, MASS, ggExtra |
| Suggests: | testthat, knitr, rmarkdown, GGally |
| Published: | 2016-10-03 |
| Author: | Jeremy Oakley |
| Maintainer: | Jeremy Oakley <j.oakley at sheffield.ac.uk> |
| BugReports: | https://github.com/OakleyJ/SHELF/issues |
| License: | GPL-2 | GPL-3 |
| URL: | https://github.com/OakleyJ/SHELF |
| NeedsCompilation: | no |
| Materials: | NEWS |
| CRAN checks: | SHELF results |
| Reference manual: | SHELF.pdf |
| Vignettes: |
Eliciting a Dirichlet Distribution Eliciting Dependent Distributions using Multivariate Normal Copulas Getting Started with SHELF |
| Package source: | SHELF_1.2.1.tar.gz |
| Windows binaries: | r-devel: SHELF_1.2.1.zip, r-release: SHELF_1.2.1.zip, r-oldrel: SHELF_1.1.0.zip |
| OS X Mavericks binaries: | r-release: SHELF_1.1.0.tgz, r-oldrel: SHELF_1.1.0.tgz |
| Old sources: | SHELF archive |
Please use the canonical form https://CRAN.R-project.org/package=SHELF to link to this page.