Retrieves the number of chunks used by an HDF5 dataset.
H5Dget_num_chunks(h5dataset)An object of class H5IdComponent representing the dataset from which chunks will be counted.
An integer value indicating the number of chunks present in the dataset or selected region.
Note, this function only returns the number of chunks that actually have data written to them.
It does not return the theoretical number of chunks in a dataset or intersection with a
dataspace. For example, if an empty dataset is created and but no values have been written
to it H5Dget_num_chunks() will return 0. This can be seen in the examples below.
The C API also provides an optional parameter to constrain the query by providing a dataspace selection. However this argument is not currently used at the C level and so is ommitted here.
file <- tempfile(fileext = ".h5")
fid <- H5Fcreate(file)
## Create a dataset that will be represented by 4 chunks if complete
h5createDataset(file, "data", dims = c(10, 10), chunk = c(5, 5), storage.mode = "integer")
#> Warning: An open HDF5 file handle exists. If the file has changed on disk meanwhile, the function may not work properly. Run 'h5closeAll()' to close all open HDF5 object handles.
did <- H5Dopen(fid, "data")
## Here we return 0 chunks as no values have been written
H5Dget_num_chunks(did)
#> [1] 0
## Now write data to half the dataset
h5writeDataset(obj = matrix(1:50, nrow = 10), h5loc = fid, name = "/data", index = list(1:10, 1:5))
## We now see it contains 2 chunks
H5Dget_num_chunks(did)
#> [1] 2
## Now write the complete dataset, overwriting the existing values
h5writeDataset(obj = matrix(201:300, nrow = 10), h5loc = fid, name = "/data", index = NULL)
## We now see it contains 4 chunks
H5Dget_num_chunks(did)
#> [1] 4
## Tidy up op handles
h5closeAll(did, fid)