Description Usage Arguments Details Value Author(s) References See Also Examples
This function assesses statistical significance of a wcr
or wnet
fit by referring the cross-validation criterion to a permutation distribution.
1 2 3 4 5 6 7 8 | wcr.perm(y, xfuncs, min.scale = 0, nfeatures, ncomp, method = c("pcr", "pls"),
covt = NULL, nrep = 1, nsplit = 1, nfold = 5, nperm = 20,
perm.method = NULL, family = "gaussian",
seed.real = NULL, seed.perm = NULL, ...)
wnet.perm(y, xfuncs, min.scale = 0, nfeatures = NULL, alpha = 1, lambda,
covt = NULL, nrep = 1, nsplit = 1, nfold = 5, nperm = 20,
perm.method = NULL, family = "gaussian",
seed.real = NULL, seed.perm = NULL, ...)
|
y, xfuncs, min.scale, nfeatures, method, covt, family, nsplit, nfold |
arguments passed to |
ncomp |
number of components; passed to |
alpha, lambda |
tuning parameters, passed to |
nrep |
number of replicates to average over, when computing the real-data CV criterion. |
nperm |
number of permutations. The default is suitable for toy applications only. |
perm.method |
either NULL or one of
See Details. |
seed.real |
the seed for random data division for real data. If |
seed.perm |
the seed for random data division for permuted data. If |
... |
other arguments passed to |
Permutation tests of this type are discussed, in a classification setting, by Ojala and Garriga (2010). Permuting the responses (perm.method="responses"
) is appropriate when regressing on functions/images only, with no scalar covariates. For linear regression with covariates, it is preferable to first regress on the covariates, and then permute the residuals (perm.method="y.residuals"
). For logistic regression this is not feasible; but, following Potter (2005), one can instead permute the residuals from a regression of the functions/images on the covariates (perm.method="x.residuals"
). When perm.method=NULL
(the default), "responses"
is used if covt
is NULL
, and "x.residuals"
otherwise.
cv |
CV value for the real data (averaged over |
cv.perm |
CV values for the permuted data. |
pvalue |
p-value for the permutation test. |
Lan Huo lan.huo@nyumc.org
Ojala, M., and Garriga, G. C. (2010). Permutation tests for studying classifier performance. Journal of Machine Learning Research, 11, 1833–1863.
Potter, D. M. (2005). A permutation test for inference in logistic regression with small- and moderate-sized data sets. Statistics in Medicine, 24, 693–708.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 | n = 200; d = 64
# Create true coefficient function
ftrue = matrix(0,d,d)
ftrue[40:46,34:38] = 1
# Generate random functional predictors, and scalar responses
ii = array(rnorm(n*d^2), dim=c(n,d,d))
iimat = ii; dim(iimat) = c(n,d^2)
yy = iimat %*% as.vector(ftrue) + rnorm(n, sd=.3)
obj.wnet.perm <- wnet.perm(yy, xfuncs = ii, min.scale = 4, nfeatures = 200, alpha = 1,
nperm = 10)
obj.wcr.perm <- wcr.perm(yy, xfuncs = ii, min.scale = 4, nfeatures = 20, ncomp = 6,
cv1 = TRUE, method = "pls", nperm = 10)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.