usage survey

[mschubert]

This is an overview of how narray is used in R packages present on CRAN, to get some idea where to focus future development. If anyone reading this uses narray themselves, please add as a comment.

• audrex - split array
• clustermq - split array
• codez - split array
• dymo - split array with purrr::map, could probably use narray::map
• FunCC - rep for vector to matrix
• janus - subset, probably doable by [
• jenga - split, mostly with purrr::map
• lambdaTS - split, mostly with purrr::map2; provides smart_split, shouldn't this be just drop=TRUE?
• proteus - smart_split as above
• segen - split
• tetragon - split

Code outside packages:

• ClimateImpactLab is using stack

Other packages providing similar functionality:

• arrayhelpers - some common functions like apply and grouped version, can't be used much because e.g. still call to browser() in the code
• arrApply - apply for commonly used functions in Rcpp, similar to MarixStats
• implicitExpansion - broadcasting for R
• keep - own array type to not drop dims
• lazyarray - on-disk array, will be replaced by filearray
• listArray - key-based dimension access
• listarrays - functional programming with arrays: binding, splitting, setting dimnames, etc.
• RcppArray - creation of type-specific arrays with Rcpp
• tensor - tensor products (single R function)
• via - not sure, vignettes largely empty