Changes to version 1.7-2
o User level changes
- Provided new dcor.test function, similar to dcov.test but using the
distance correlation as the test statistic.
- Number of replicates R for Monte Carlo and permutation tests now matches
the argument of the boot::boot function (no default value, user must specify).
- If user runs a test with 0 replicates, p-value printed is NA
o Internal changes
- energy_init.c added for registering routines
Changes to version 1.7-0
o Partial Distance Correlation statistics and tests added
- pdcov, pdcor, pdcov.test, pdcor.test
- dcovU: unbiased estimator of distance covariance
- bcdcor: bias corrected distance correlation
- Ucenter, Dcenter, U_center, D_center: double-centering and U-centering utilities
- U_product: inner product in U-centered Hilbert space
o updated NAMESPACE and DESCRIPTION imports, etc.
o revised package Title and Description in DESCRIPTION
o package now links to Rcpp
o mvnorm c code ported to c++ (mvnorm.cpp); corresponding changes in Emvnorm.R
o syntax for bcdcor: "distance" argument removed, now argument can optionally
be a dist object
o syntax for energy.hclust: first argument must now be a dist object
o default number of replicates R in tests: for all tests, R now defaults to 0
or R has no default value.
Changes to version 1.6.2
o inserted GetRNGstate() .. PutRNGState around repl.
loop in dcov.c.
Changes to Version 1.6.1
o replace Depends with Imports in DESCRIPTION file
Changes to Version 1.6.0
o implementation of high-dim distance correlation t-test
introduced in JMVA Volume 117, pp. 193-213 (2013).
o new functions dcor.t, dcor.ttest in dcorT.R
o minor changes to tidy other code in dcov.R
o removed unused internal function .dcov.test
Changes to Version 1.5.0
o NAMESPACE: insert UseDynLib; remove zzz.R, .First.Lib()
Changes to Version 1.4-0
o NAMESPACE added.
o (dcov.c, Eindep.c) Unused N was removed.
o (dcov.c) In case dcov=0, bypass the unnecessary loop
that generates replicates (in dCOVtest and dCovTest).
In this case dcor=0 and test is not significant.
(dcov=0 if one of the samples is constant.)
o (Eqdist.R) in eqdist.e and eqdist.etest, method="disco"
is replaced by two options: "discoB" (between sample
components) and "discoF" (disco F ratio).
o (disco.R) Added disco.between and internal functions
that compute the disco between-sample component and
corresponding test.
o (utilities.c) In permute function replaced rand_unif
with runif.
o (energy.c) In ksampleEtest the pval computation
changed from ek/B to (ek+1)/(B+1) as it should be for
a permutation test, and unneeded int* n removed.
Changes to Version 1.3-0
o In distance correlation, distance covariance functions
(dcov, dcor, DCOR) and dcov.test, arguments x and y can now
optionally be distance objects (result of dist function or
as.dist). Matrices x and y will always be treated as data.
o Functions in dcov.c and utilities.c were modified to support
arguments that are distances rather than data. In utilities.c
the index_distance function changed. In dcov.c there are many
changes. Most importantly for the exported objects, there is
now an extra required parameter in the dims argument passed
from R. In dCOVtest dims must be a vector c(n, p, q, dst, R)
where n is sample size, p and q are dimensions of x and y,
dst is logical (TRUE if distances) and R is number of replicates.
For dCOV dims must be c(n, p, q, dst).
Changes to Version 1.2-0
o disco (distance components) added for one-way layout.
o A method argument was added to ksample.e, eqdist.e, and
eqdist.etest, method = c("original", "disco").
o A method argument was added to edist, which summarizes cluster
distances in a table:
method = c("cluster","discoB","discoF"))