Products
ironArray
For Python developers looking for a high-performance library to perform computations on compressed multidimensional arrays, or slices of them, while consuming exceptionally low resources.
ironArray Community
For Python developers looking for a simple yet performant library to slice and dice compressed multidimensional arrays in an effective way.
We have prepared a Jupyter environment for you, where you can play with our tutorials or test your use-case with ironArray!
|
Features
|
ironArray Community
|
ironArray
|
|---|---|---|
|
Multidimensional arrays
|
|
|
|
Transparent data compression
|
|
|
|
Metadata support
|
|
|
|
Multiple compressors and filters
|
|
|
|
Store in memory or on disk
|
|
|
|
Efficient storage for sparse data
|
|
|
|
Powerful constructors
|
|
|
|
Slicing
|
|
|
|
Views
|
|
|
|
Just-In-Time compiler for computational kernels
|
|
|
|
Custom Python kernels
|
|
|
|
Mathematical expression kernels
|
|
|
|
Fast constructors for random distributions
|
|
|
|
Computation optimisation via Artificial Intelligence
|
|
|
|
Efficient reductions
|
|
|
Comparing Zarr/HDF5 with ironArray Community
|
Features
|
Zarr
|
HDF5
|
ironArray Community
|
|---|---|---|---|
|
Zarr/HDF5 API compatibility
|
|
|
|
|
Multidimensional arrays
|
|
|
|
|
Transparent data compression
|
|
|
|
|
Transparent metadata compression
|
|
|
|
|
Compression library
|
numcodecs/Blosc1
|
hdf5plugin/Blosc1
|
Blosc2
|
|
Fast slicing
|
|
|
|
|
Multi-file (sparse) storage
|
|
|
|
|
Single file (contiguous) storage
|
|
|
|
|
Open data format
|
|
|
|
Comparing Dask & Zarr with ironArray
|
Features
|
Dask & Zarr
|
ironArray
|
|---|---|---|
|
Computations on multidimensional arrays
|
|
|
|
Fast and efficient reductions
|
|
|
|
Fast expression evaluation
|
|
|
|
Compiled User Defined Functions
|
|
|
|
Low memory (RAM) consumption
|
|
|
|
Open Data Format
|
|
|
|
Distributed computation
|
|
Planned
|
