NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI · #900 most downloaded on PyPI
Path optimization of einsum functions.
Last release 2 years ago
no release in 18 months
Ships unpredictably
gaps range from 8 days to 4.2 years
Nearly every release is documented
notes for 14 of 15 stable releases
Nothing withdrawn
no release was ever pulled
10 years old
15 releases · first in 2016
NumPy has been removed from opt_einsum as a dependency allowing for more flexible installs. Type hints have been added to the code base as well as an
NumPy has been removed from opt_einsum as a dependency allowing for more flexible installs. Type hints have been added to the code base as well as an overhaul of the documentation to MkDocs here: https://dgasmith.github.io/opt_einsum/
New Features
Enhancements
memory_limit was set very low for the dp algorithm.backend=None which is equivalent to backend='auto'ConfigParser insetad of SafeConfigParser for Python 3.12 compatability.backend='jaxlib' is now an alias for the jax libraryruff for formatting and linting.numpy-specific keyword args from being explicitly defined in contract and uses **kwargs instead.Bug Fixes
dp would not work for scalar-only contractions.parse_einsum_input would not correctly respect shape-only contractions.parse_einsum_input where an output subscript specified multiple times was not correctly caught.PathInfo would cause an error.One column per quarter.
Adds a object backend for optimized contractions on arbitrary Python objects.
Adds a object backend for optimized contractions on arbitrary Python objects.
object based backend so that contract(backend='object') can be used on arbitrary objects such as SymPy symbols.contract backend cannot be found.remaining category for the contract_path output to only show up to 20 tensors to prevent issues with the quadratically scaling memory requirements and the number of print lines for large contractions.( #131 ) Fixes an einstein subscript error message to point to the correct value.
Small fixes for the dp path and support for a new mars backend.
Adds a new dynamic programming algorithm to the suite of paths.
Adds a new dynamic programming algorithm to the suite of paths.
Alters setup.py to correctly state that opt_einsum requires Python 3.5+. This will now correctly effect PyPI and other pip mirror downloads, v3.0.0 wi
Alters setup.py to correctly state that opt_einsum requires Python 3.5+. This will now correctly effect PyPI and other pip mirror downloads, v3.0.0 will be removed from PyPI to prevent further issues.
( #77 ) Fixes a PyTorch v1.0 JIT tensor shape issue.
Minor tweak to release procedure.
(#48) Intermediates can now be shared between contractions, see here for more details.
New features:
Enhancements:
Bug fixes:
a-z subset valid for PyTorch's einsum implementation.Fixes unicode issue for large numbers of tensors in Python 2.7.
Bug fixes:
README.md.Ensures versioneer.py is in MANIFEST.in for a clean pip install.
Bug Fixes:
versioneer.py is in MANIFEST.in for a clean pip install.Minor tweak to release procedure.
Bug Fixes:
opt_einsum is a powerful tensor contraction order optimizer for NumPy and related ecosystems.
opt_einsum is a powerful tensor contraction order optimizer for NumPy and related ecosystems.
greedy order optimization algorithm has been tuned to be able to handle hundreds of tensors in several seconds.Nothing published for this version
A large step towards to a full 1.0 release. BLAS usage is now automatically applied to all operations. Future releases will be more careful with regar
A large step towards to a full 1.0 release. BLAS usage is now automatically applied to all operations. Future releases will be more careful with regard to views and needless data copying.
Your coding agent can read these notes before it upgrades. Set up the MCP server →