confinterpret
Descriptive Interpretations of Confidence Intervals
Produces descriptive interpretations of confidence intervals. Includes (extensible) support for various test types, specified as sets of interpretations dependent on where the lower and upper confidence limits sit. Provides plotting functions for graphical display of interpretations.
An R package for interpreting confidence intervals.
The goal of confinterpret is to allow users to produce standardised descriptive interpretations from confidence intervals in R. The package supports researchers to focus more on generating estimates with associated ranges of uncertainty, rather than on single point-estimates, such as are common when conclusions are drawn from p-values.
The package includes (extensible) support for various test types, which are specified as sets of interpretations dependent on where the lower and upper confidence limits sit.
A collection of interpretation_set objects are provided in the package, to cover certain common situations where the confidence interval is of an effect size comparing two groups. Convenience / wrapper functions are provided for using each of these:
- Superiority tests:
interpret_superiority() - Non-inferiority tests:
interpret_noninferiority() - Equivalence tests:
interpret_equivalence()
Installation
confinterpret is on CRAN, so you can install it in the normal way:
install.packages("confinterpret")You can also install the development version of confinterpret from github with:
# install.packages("devtools")devtools::install_github("jimvine/confinterpret")Example
This is a basic example which shows you how to solve a common problem:
# Set up a dummy confidence intervalci_above <- matrix(c(0.15, 0.25), nrow = 1, dimnames = list("estimate", c("2.5 %","97.5 %"))) ci_above_sup <- interpret_superiority(ci_above) # Access the short-form interpretation:ci_above_sup$interpretation_short # Access the full interpretation:ci_above_sup$interpretation # Access the full interpretation, marked up with markdown:ci_above_sup$interpretation_md # Plot a chart of the interpretation (specify a nice colour-scheme first):grDevices::palette(c("#FF671F99", "#F2A90099", "#0085CA99"))plot(ci_above_sup)News
confinterpret 1.0.0
S3 class for interpretation_result
- The output of
confinterpret()now returns an object of class `interpretation_result. - These results now include the
interpretation_setobject that were used to generate them (previous versions just included the name of theinterpretation_set).
Plotting
- Added plot method for
interpretation_resultobjects. Plots the result on a chart, using theinterpretation_setand associated parameters to draw a relevant canvas that shows the boundaries etc. - Added a plotting function that works for a collection of
interpretation_resultobjects. See?plot_interpretation_result_list. - All plotting functions support (and default to) drawing the intervals as curves rather than simple boxes. (Boxes can still be produced by specifying
interval_type = "unif".)
Printing
- Print method provided for
interpretation_setobjects. - Print method provided for
interpretation_resultobjects.
confinterpret 0.2.0
Plotting
- Improved
plot.interpretation_set()function - now takes account of length ofextra_boundariesnames when setting the bottom margin.
Documentation
- Corrections to help files (including errors in which
interpretation_setsome of the functions use). - Added examples to help files for
interpret_superiority(),interpret_noninferiority(), andinterpret_equivalence(). - Added documentation on placeholders in supplied
interpretation_setobjects. - Added Value section to help files for
interpret_superiority(),interpret_noninferiority(), andinterpret_equivalence().
confinterpret 0.1.0
NB: v0.1.0 was accepted on CRAN (2017-03-15): https://CRAN.R-project.org/package=confinterpret
- Initial version.
- Added a
NEWS.mdfile to track changes to the package.