In Mkossmeier/metaviz: Forest Plots, Funnel Plots, and Visual Funnel Plot Inference for Meta-Analysis

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "README-"
)

Package metaviz

A compilation of functions to plot meta-analytic data using 'ggplot2'. Currently allows to create forest plots, funnel plots, and many of their variants, such as rainforest plots, thick forest plots, additional evidence contour funnel plots, and sunset funnel plots. In addition, functionalities for visual inference with the funnel plot are provided.

Contact

Questions, ideas, criticism: michael.kossmeier@univie.ac.at.

History

Package metaviz version 0.1.0

February 6, 2017 first release on CRAN: https://CRAN.R-project.org/package=metaviz

Package metaviz version 0.1.1 (not yet on CRAN)

March 14, 2017: Added a new function for visual inference with funnel plots: funnelinf. Also available as shiny app: https://metaviz.shinyapps.io/funnelinf_app/

Package metaviz version 0.1.1

June 29, 2017: Version 0.1.1 submitted to CRAN

Package metaviz version 0.2

March 16, 2018: Version 0.2 submitted to CRAN
Greatly extended functionalities to create different types of forest plots and to align tables with study-level or summary-level information (viz_forest)
Added numerous funnel plot variants including additional evidence contour funnel plots with the function viz_funnel.

Package metaviz version 0.3

January 14, 2019: Version 0.3 submitted to CRAN
Added a dedicated function for sunset (power-enhanced) funnel plots (viz_sunset). Also available as shiny app: https://metaviz.shinyapps.io/sunset/

Package metaviz version 0.3.1

April 7th, 2020: Version 0.3.1 submitted to CRAN
Fixed problem of up-side-down funnel plots (viz_funnel, viz_sunset, funnelinf) with y_axis = "se", which seemed to occur with newer versions of ggplot2 installed.
Function viz_funnel now allows to customize funnel plot contours for random effects models via the contours_type argument.
Changed behavior of viz_forest, viz_funnel and viz_sunset, such that when output of function rma.uni from package metafor is used as input, then the method argument is now extracted from the rma.uni object.