The 'bigMatrix' class is no longer implemented in the NetRep
package: the shared memory approach was incompatabile with high performance
compute clusters, so the parallel permutation procedure has been translated
into C++ code (which is also much faster). The disk.matrix
class should now be used instead when analysing large datasets.
load.bigMatrix(backingfile)path to the backingfile for the 'bigMatrix'. The
file extension must be omitted.
This function will convert 'bigMatrix' data saved by previous
versions of NetRep to a serialized R matrix saved in the same location
and return a disk.matrix object with the associated file path.
If this conversion has taken place already the function will throw a warning.
This function will also convert the 'bigMatrix' descriptor file to a
big.matrix descriptor file to preserve
compatability with functions in the bigmemory package. If this
functionality is not required, the files with the extensions ".bin" and
".desc" may be removed.
A note for users using multi-node high performance clusters:
'big.matrix' objects are not suitable for general usage. Access
to file-backed shared memory segments on multi-node systems is very slow
due to consistency checks performed by the operating system. This becomes
exponentially worse the more R sessions there are simultaneously accessing
the shared memory segment, e.g. through parallel foreach loops.