Repository files navigation

trending aims to provides a coherent interface to several modellingtools. Whilst it is useful in an interactive context, it’s main focus isto provide an intuitive interface on which other packages can bedeveloped(e.g. trendbreaker).

You can install the stable version fromCRAN with:

install.packages("trending")

The development version can be installed fromGitHub with:

if (!require(remotes)) {  install.packages("remotes")}remotes::install_github("reconverse/trending",build_vignettes=TRUE)

• Model specification: Interfaces to common models through intuitivefunctions;lm_model(),glm_model(),glm_nb_model andbrms_model*.

• Model fitting and prediction: Once specified, models can be fit todata and generate confidence and prediction intervals for future datausingfit() andpredict().

• Error and warning catching: The provided methods forfit andpredict catch all warnings and errors, returning the output andthese captured values in a list.

*   Requiresbrms

An overview oftrending is provided in the included vignette: *vignette("Introduction", package = "trending")

Bug reports and feature requests should be posted ongithub using theissue system. Allother questions should be posted on theRECON slack channel seehttps://www.repidemicsconsortium.org/forum/ for details on how tojoin.

• Gavin Simpson; Our method to calculate prediction intervals followsone that he described in two posts on hisblog; seepart1andpart2.

• John Haman and Matthew Avery; Our implementation of predictionintervals was guided by their bootstrapped approach within theciTools package.