Getting started with wbcmd • wbcmd

wbcmd wraps the Connectome Workbench wb_command CLI for use in R. Every function calls the same central dispatcher, so your environment only needs to be configured once.

Prerequisites

Install Connectome Workbench from the HCP website. Then load the package and verify your setup:

library(wbcmd)
wb_sitrep()

If the binary is not detected automatically, point wbcmd to it:

set_wb_path("/opt/workbench/bin_linux64/wb_command")

You can also set the path via environment variable or R option. The lookup order is:

R option wbcmd.path
Environment variable WB_PATH
Platform-specific defaults
System PATH

Running commands

The wb_cmd() function is the universal entry point. Pass the subcommand name and arguments exactly as you would on the command line:

wb_cmd("-version")
wb_cmd("-file-information", c("data.dscalar.nii"))

Convenience wrappers

Typed wrappers handle argument construction and file validation for you.

Separate a CIFTI file

wb_cifti_separate(
  "data.dscalar.nii",
  direction = "COLUMN",
  metric = list(
    CORTEX_LEFT = "left.func.gii",
    CORTEX_RIGHT = "right.func.gii"
  )
)

Smooth CIFTI data

wb_cifti_smoothing(
  "data.dtseries.nii",
  surface_sigma = 4,
  volume_sigma = 4,
  direction = "COLUMN",
  cifti_out = "smoothed.dtseries.nii",
  left_surface = "left.midthickness.surf.gii",
  right_surface = "right.midthickness.surf.gii"
)

Parcellate

wb_cifti_parcellate(
  "data.dtseries.nii",
  "atlas.dlabel.nii",
  direction = "COLUMN",
  cifti_out = "parcellated.ptseries.nii"
)

Resample a surface

wb_surface_resample(
  "lh.pial.surf.gii",
  "lh.sphere.32k.surf.gii",
  "lh.sphere.164k.surf.gii",
  method = "BARYCENTRIC",
  surface_out = "lh.pial.164k.surf.gii"
)

Controlling verbosity

By default wbcmd prints each command before running it. Silence this with:

options(wbcmd.verbose = FALSE)

Or set the WB_VERBOSE environment variable to FALSE.