Hierarchical inference testing (HIT) for (generalized) linear models with correlated covariates applicable to high-dimensional settings.
The current built and test status for Linux (Mac)
and for Windows
.
Hierarchical inference testing (HIT) for (generalized) linear models with correlated covariates. HIT is furthermore applicable to high-dimensional settings. For details see:
Mandozzi, J. and Buehlmann, P. (2015). Hierarchical testing in the high-dimensional setting with correlated variables. Journal of the American Statistical Association. Preprint
Klasen, J. R. et al. (2016). A multi-marker association method for genome-wide association studies without the need for population structure correction. Nature Communications. Paper
The package can be installed from CRAN,
install.packages("hit")
or via
devtools
, if you haven't devtools
installed yet you have to do so first.
# install.packages("devtools")devtools::install_github("QTCAT/hit")
The hit
-function example gives an overview of the functionality of the
package and can be accessed once the package is loaded.
library(hit)example(hit)
Changes:
hit
: second path for lambda estimation, with a new lambda for each sample split.Changes:
reorder
: a more general reorder function for hierarchyChanges:
hit
: remove the alpha optimization and allow only one alpha value
hit
: allow the response variable (y) to be Poisson distributed
summary.hit
: bug fix
Changes:
hit
: bug fixes in summary and some other small fixesChanges:
hit
: sticks to cross-validation as selection method
hit
: arguments have changed
Changes:
fast.anova
: a new fast.glmanova
method for GLM's
hit
: now is able to deal with binomial responses
hit
: a new selection method for the active set