Performance tuning options for rxDForest/rxDTree

Performance tuning options for rxDForest/rxDTree
PerformancexVal              = 0      # (rxDTree) number of cross-validations during fitting

computeOobError   = -1     # (rxDForest) do not compute OOB prediction errors

scheduleOnce      = TRUE   # launch via rxExec (if modest data; best for Hadoop)

                           # Requires nTree = #trees / #tasks

blocksPerRead     = large  # if the input is an XDF file – Warning: reduce this if memory consumption is an issue

verbose           = 0      # minimize console output

reportProgress    = 0      # ditto

useSparseCube     = FALSE  # sparse cube is a bit slower so avoid if possible
Note This is a "FAST PUBLISH" article created directly from within the Microsoft support organization. The information contained herein is provided as-is in response to emerging issues. As a result of the speed in making it available, the materials may include typographical errors and may be revised at any time without notice. See Terms of Use for other considerations.

Article ID: 3104235 - Last Review: 10/29/2015 06:56:00 - Revision: 1.0

Revolution Analytics

  • KB3104235