SVDWorker()SVDMaster.RdSVDMaster objects instantiate and run a distributed SVD computation
SVDWorker() which goes hand-in-hand with this object
new()SVDMaster objects instantiate and run a distributed SVD computation
SVDMaster$new(defn, debug = FALSE)
defna computation definition
debuga flag for debugging, default FALSE
R6 SVDMaster object
kosher()Check if inputs and state of object are sane. For future use
SVDMaster$kosher()
TRUE or FALSE
updateV()Return an updated value for the V vector, normalized by arg
SVDMaster$updateV(arg)
argthe normalizing value
...other args ignored
updated V
updateU()Update U and return the updated norm of U
SVDMaster$updateU(arg)
argthe normalizing value
...other args ignored
updated norm of U
fixFit()Construct the residual matrix using given the V vector and d so far
SVDMaster$fixFit(v, d)
vthe value for v
dthe value for d
result
reset()Reset the computation state by initializing work matrix and set up starting values for iterating
SVDMaster$reset()
addSite()Add a url or worker object for a site for participating in the distributed computation. The worker object can be used to avoid complications in debugging remote calls during prototyping.
SVDMaster$addSite(name, url = NULL, worker = NULL)
nameof the site
urlweb url of the site; exactly one of url or worker should be specified
workerworker object for the site; exactly one of url or worker should be specified
run()Run the distributed Cox model fit and return the estimates
SVDMaster$run(thr = 1e-08, max.iter = 100)
thrthe threshold for convergence, default 1e-8
max.iterthe maximum number of iterations, default 100
a named list of V, d
summary()Return the summary result
SVDMaster$summary()
a named list of V, d
clone()The objects of this class are cloneable with this method.
SVDMaster$clone(deep = FALSE)
deepWhether to make a deep clone.