Apply a function across all locations in each parcel, for a pair of data and
parcellation "xifti" objects that are in registration with one
another. By default, the mean value in each parcel is calculated.
apply_parc(
xii,
parc,
FUN = mean,
mwall_value = NA,
return_as = c("matrix", "xifti"),
...
)A \(P \times Q\) matrix, where \(P\) is the number of parcels and
\(Q\) is the length of the output of FUN. (For mean,
\(Q = 1\)).
The "xifti" data to apply the function over, within each
parcel.
The "xifti" "dlabel" parcellation. Each parcel is defined
by a unique key in the label table. If there are multiple columns, only the
first column will be used. Alternatively, parc can just be a vector
of keys whose length is the number of data locations in "xii".
A function that takes as input an \(M \times N\) matrix (\(M\)
locations in a given parcel, and \(N\) measurements/columns in xii)
and outputs a constant-sized (\(Q\)) numeric vector. Default: mean.
Use colMeans to obtain the average timeseries of each parcel, such as
in order to compute functional connectivity.
If there is a medial wall in xii, what should value
should medial wall locations be replaced with prior to calculation?
Default: NA.
"matrix" (default) where each row corresponds to a
parcel, or a "xifti" object where each location's value is the value
of its corresponding parcel?
Additional arguments to FUN, e.g. na.rm=TRUE.
Ignored if FUN=="quick_mean".
Other parcellation-related:
load_parc(),
load_sub_parc(),
parc_add_subcortex(),
parc_borders(),
parc_vals_to_xifti()
Other manipulating xifti:
add_surf(),
apply_xifti(),
combine_xifti(),
convert_xifti(),
impute_xifti(),
merge_xifti(),
move_to_mwall(),
move_to_submask(),
newdata_xifti(),
remap_cifti(),
remove_xifti(),
resample_cifti(),
resample_cifti_from_template(),
scale_xifti(),
select_xifti(),
set_names_xifti(),
smooth_cifti(),
transform_xifti()