The surveyverse package makes it easy to install and load a collection of R packages for analyzing complex survey data. The key package in this collection is the survey package of Lumley (2004). The srvyr package of Freedman Ellis and Schneider (2016) extends survey by allowing users to write code for survey data analysis in the style of the tidyverse. The svrep package extends survey’s ability to create and use replicate weights based on methods such as the generalized bootstrap or successive difference replication.
Users should cite the specific packages in this collection rather
than citing surveyverse. Citations for individual packages can be
obtained using the citation() function, for example by
calling citation('survey').
You can install surveyverse from CRAN like so:
install.packages('surveyverse')Calling library(surveyverse) will load the core set of
packages in the surveyverse collection.
library(surveyverse)
#> ── Attaching core surveyverse packages ──────────────────── surveyverse 0.1.1 ──
#> ✔ survey 4.5 ✔ svrep 0.9.1
#> ✔ srvyr 1.3.1
#> ── Conflicts ──────────────────────────────────────── surveyverse_conflicts() ──
#> ✖ survey::dotchart() masks graphics::dotchart()
#> ✖ srvyr::filter() masks stats::filter()
#> ℹ Use the conflicted package to force all conflicts to become errors
#> ℹ Please cite specific packages rather than surveyverse; see `citation('surveyverse')`Loading the package in this way will display any namespace conflicts with other packages that are currently loaded.
The set of packages in the collection can be viewed by calling
surveyverse_packages(). The following non-core packages are
installed with surveyverse but are not loaded:
svylme: For fitting linear mixed-effects models to complex survey data
svyVGAM: For fitting vector generalised linear models, i.e., models supported by the VGAM package
The full list of core and non-core packages can be viewed by calling
surveyverse_packages(core_only = FALSE).