Inspect HDF5 Dataset Creation Properties

Retrieves the Dataset Creation Property List (DCPL) details including storage layout, chunk dimensions, and a detailed list of all applied filters.

Usage

h5_inspect(file, name)

Arguments

file: The path to the HDF5 file.
name: The full path of the dataset to inspect.

Value

An object of class inspect (a named list) containing:

layout: A string indicating storage layout (e.g., "chunked", "contiguous").
chunk_dims: A numeric vector of chunk dimensions, or NULL if not chunked.
filters: A list describing each filter applied.

Details

ZFP can't compress chunks with no dimension longer than one (such as a single value) or more than four. h5_write() uses the default pipeline for them, but in files from other software ZFP may have declined such a dataset: every chunk is stored uncompressed, and print() shows the filter as zfp (declined: stored uncompressed). With hdf5lib 2.2.0.1 or later, Blosc2's ZFP codecs likewise decline chunks of fewer than four values, shown as, e.g., blosc2 [zfp-acc] (declined: stored uncompressed).

Examples

file <- tempfile(fileext = ".h5")

compress <- h5_compression('lz4-9', int_packing = TRUE, checksum = TRUE)
h5_write(matrix(5001:5100, 10, 10), file, "packed_mtx", compress = compress)
h5_inspect(file, "packed_mtx")
#> <HDF5 Dataset Properties>
#>   Type:    uint16              Size:    200.00 B
#>   Layout:  chunked             Disk:    120.00 B
#>   Chunks:  [10 x 10]           Ratio:   1.67x
#>   Pipeline: scaleoffset -> lz4 -> fletcher32 

mtx <- matrix(rnorm(1000), 100, 10)
h5_write(mtx, file, "float_mtx", compress = 'blosc2-zfp-prec-3')
res <- h5_inspect(file, "float_mtx")
print(res)
#> <HDF5 Dataset Properties>
#>   Type:    float64             Size:    7.81 KB
#>   Layout:  chunked             Disk:    1.48 KB
#>   Chunks:  [100 x 10]          Ratio:   5.29x
#>   Pipeline: blosc2 [zfp-prec] 

# Print the raw cd_values for blosc2
dput(res$filters[[1]]$cd_values)
#> c(2987393793, 0, 8, 8000, 5, 0, 34, 2, 100, 10)

unlink(file)