Bitshuffle is an algorithm that rearranges typed, binary data for
improving compression, as well as a python/C package that implements
this algorithm within the Numpy framework.
.
The library can be used along side HDF5 to compress and decompress
datasets and is integrated through the dynamically loaded filters
framework. Bitshuffle is HDF5 filter number 32008.
.
Algorithmically, Bitshuffle is closely related to HDF5's Shuffle
filter except it operates at the bit level instead of the byte level.
Arranging a typed data array in to a matrix with the elements as the
rows and the bits within the elements as the columns, Bitshuffle
"transposes" the matrix, such that all the least-significant-bits
are in a row, etc. This transpose is performed within blocks of
data roughly 8kB long.
.
This does not in itself compress data, only rearranges it for more
efficient compression. To perform the actual compression you will
need a compression library. Bitshuffle has been designed to be well
matched Marc Lehmann's LZF as well as LZ4. Note that because
Bitshuffle modifies the data at the bit level, sophisticated entropy
reducing compression libraries such as GZIP and BZIP are unlikely to
achieve significantly better compression than simpler and faster
duplicate-string-elimination algorithms such as LZF and LZ4.
Bitshuffle thus includes routines (and HDF5 filter options) to apply
LZ4 compression to each block after shuffling.
.
The Bitshuffle algorithm relies on neighbouring elements of a dataset
being highly correlated to improve data compression. Any correlations
that span at least 24 elements of the dataset may be exploited to
improve compression.
