In-memory OLAP cubes R data type. Uses high performance C-implemented data.table R package.
- scalable multidimensional hierarchical
array
alternative - uses data.table under the hood
- use base R
array
query API-
[.data.cube
uses base R[.array
method API for slice and dice of dimension keys
-
- extends base R
array
query API- slice and dice on dimension attributes of various levels in hierarchy with
.(time_year = 2011:2013)
- aggregate by collapse dimensions with
`-`()
, also with pre-filtering - rollup and cube over provided groupings with
`+`()
for rollup and`^`()
for cube
- slice and dice on dimension attributes of various levels in hierarchy with
-
apply.data.cube
uses base Rapply
like API -
rollup
fordata.cube
- bind grouping dimension to retain cube normalization and avoid double counting
- for pivot use
format
/as.data.table
withdcast.data.table
API - direct access to data.cube child classes and attributes
- query optimization
- use blazingly fast data.table indices
- use data.table#1377 grouping sets
- works on sharded distributed engine using big.data.table
install.packages("data.cube", repos = paste0("https://", c(
"jangorecki.gitlab.io/data.cube",
"cloud.r-project.org"
)))
Read manual and check Subset and aggregate multidimensional data with data.cube vignette.
j.gorecki@wit.edu.pl