Run sparse CCA on a pair of matrices using ANTsR
x.SD2RES(x, z, x.test = NULL, z.test = NULL, k = 4, inmask = c(NA,
NA), sparseness = c(0.01, 1), scale.first.x = FALSE,
scale.first.center.x = FALSE, scale.first.z = FALSE,
scale.first.center.z = FALSE, resampler = "strat.boot", n.res = 4,
stratify.var = NULL, train.p = 0.75, strat.n.bins = 5,
target.length = NROW(z), its = 20, cthresh = c(0, 0), perms = 0,
uselong = 0, row.sparseness = 0, smooth = 0, robust = 0,
mycoption = 1, initializationList = list(),
initializationList2 = list(), l1 = 0.05, doseg = TRUE,
priorWeight = 0, verbose = TRUE, outdir = NULL,
save.mod = ifelse(!is.null(outdir), TRUE, FALSE))Input matrix
Input matrix
(Optional) Testing set x matrix. Will be projected on vector solution
(Optional) Testing set z matrix. Will be projected on vector solution
Integer: Number of axes to project to (i.e. number of resulting dimensions you require)
Float, length 2: Required sparseness for each matrix. Defaults to c(.01, 1)
Logical: If TRUE, scale x before decomposition
Logical: If TRUE, and scale.first.x is TRUE, also center x
Logical: See above
Resampling method to use (with resample)#' @param
Numeric vector (optional): Variable used for stratification. Defaults to y
Float (0, 1): Fraction of cases to assign to traininig set for resampler = "strat.sub"
Integer: Number of groups to use for stratification for
resampler = "strat.sub" / "strat.boot"
Integer: Number of cases for training set for resampler = "strat.boot".
Default = length(y)
Integer: Number of iterations for ANTsR decomposition
Integer, pair: Cluster threshold for ANTsR decomposition. Used for voxerlwise data Default = c(0, 0), which should be used for ROI data
Integer: Number of permutations for ANTsR decomposition
Logical, binary: If 1, enforce solutions on each side to be the same. Default = 0.
See ANTsR::sparseDecom2("uselong")
Float (0, 1]: subject / case-space sparseness
Float: Smooth the data (only if mask is provided). Default = 0
Logical, binary: If 1, Rank-transform input matrices. Default = 0
Integer 0, 1, 2: Enforce 1. spatial orthogonality, 2. low-dimensional orthogonality or 0. both Default = 1
List: Initialization for x. Default = list()
List: Initialixzation for z. Default = list()
Float: L1 normalization. Default = .05
Logical: If TRUE, orthogonalize matrix after each resample
Logical: If TRUE, print messages to screen
z Logical: See above
This is based on ANTsR::sparseDecom2boot by Brian B. Avants
Other Cross-Decomposition: x.CCA,
xdecomSelect