epoc.bootstrap(Y, U, nboots=100, bthr=NULL, method='epocG',...)
## S3 method for class 'bootsize'
plot(x, lambda.boot=NULL, B, range=c(0,1), ...)
epoc.final(epocboot, bthr=0.2, k)
Arguments
Y
mRNA, samples x genes.
U
CNA, samples x genes.
nboots
Number of bootstrap iterations to run.
method
For epoc.bootstrap method is "epocG" or "epocA".
x
A sparse network matrix or a list of the same, non-zeros are links. These come from e.g. epoc.final or epoc.bootstrap.
lambda.boot
The λs used to run the bootstrap.
B
Number of bootstrap iterations ran for 'plot.bootsize'.
range
Range of bootstrap thresholds to display.
epocboot
For epoc.final give a list of bootstraped models from epoc.bootstrap.
k
For epoc.final and epoc.bootplot select the k sparsest model.
bthr
Require presence of links in 100*bthr% of the bootstrap iterations.
...
Parameters passed down to an underlying function. For epoc.bootstrap these are passed down to "epocG" or "epocA" respectively. For epoc.bootplot and plot.bootsize parameters are passed down to the underlying plot command.
Details
epoc.bootstrap run epocA or epocG using bootstrap.
Value
epoc.bootstrap returns a list of p x p arrays of values in [0,1] where 1 is presence of link in 100% of bootstrap iterations for the k different λ values for p different genes.
epoc.final returns a sparse matrix of p x p values in [0,1] where 1 is presence of link in 100% of bootstrap iterations, but thresholded such that all values have be greater than or equal to bthr.