Reputation: 474
I'm looking to use iCAMP::bNRI.cm()
which requires a phylogenetic distance matrix as an actual matrix. However, my matrix is huge and saved as
new("big.matrix.descriptor", description = list(sharedType = "FileBacked",
filename = "pd.bin", dirname = "/dirpath/",
totalRows = 5122L, totalCols = 5122L, rowOffset = c(0, 5122
), colOffset = c(0, 5122), nrow = 5122, ncol = 5122, rowNames = NULL,
colNames = NULL, type = "double", separated = FALSE))
which I can load as a descriptor. How can I specify to iCAMP::bNRI.cm()
that my descriptor is the matrix without breaking the memory bank?
bnri.whole <- bNRI.cm(comm=otu, dis=descriptor?, nworker = 4, memo.size.GB = 50...)
Thanks!
Upvotes: 2
Views: 55