fname = "/path/to/my.gctx"
file = rhdf5::H5Fopen(fname)
data = file&"/0/DATA/0/matrix"
data[1, 1:5] # works, but coerces the matrix to a vector
data[1, 1:5, drop=FALSE] # Error: length of index has to be equal to dimensional extension of HDF5 dataset.
This is using the current Github HEAD of rhdf5.
Shouldn't the drop argument behave the same way as when subsetting a normal matrix?
Consider the following example:
This is using the current Github
HEAD
ofrhdf5
.Shouldn't the
drop
argument behave the same way as when subsetting a normal matrix?