imate

Implicit Matrix Trace Estimator. Computes the trace of functions of implicit matrices. Accelerated on CPU and CUDA-capable GPU devices.

These details have not been verified by PyPI

Project links

GitHub Statistics

View statistics for this project via Libraries.io, or by using our public dataset on Google BigQuery

Project description

imate, short for Implicit Matrix Trace Estimator, is a modular and high-performance C++/CUDA library distributed as a Python package that provides scalable randomized algorithms for the computationally expensive matrix functions in machine learning.

Install

Install with pip

pip install imate

Install with conda

conda install -c s-ameli imate

Docker Image

docker pull sameli/imate

Supported Platforms

Successful installation and tests performed on the following operating systems, architectures, and Python and PyPy versions:

Platform	Arch	Device	Python Version	PyPy Version ¹	Continuous Integration
3.9	3.10	3.11	3.12	3.8	3.9	3.10
Linux	X86-64	CPU	✔	✔	✔	✔	✔	✔	✔
	X86-64	GPU	✔	✔	✔	✔	✔	✔	✔
	AARCH-64	CPU	✔	✔	✔	✔	✔	✔	✔
	AARCH-64	GPU	✔	✔	✔	✔	✔	✔	✔
macOS	X86-64	CPU	✔	✔	✔	✔	✔	✔	✔
	X86-64	GPU ²	✖	✖	✖	✖	✖	✖	✖
	ARM-64	CPU	✔	✔	✔	✔	✖	✔	✔
	ARM-64	GPU ²	✖	✖	✖	✖	✖	✖	✖
Windows	X86-64	CPU	✔	✔	✔	✔	✖	✖	✖
Windows	X86-64	GPU	✔	✔	✔	✔	✖	✖	✖

Python wheels for imate for all supported platforms and versions in the above are available through PyPI and Anaconda Cloud. If you need imate on other platforms, architectures, and Python or PyPy versions, raise an issue on GitHub and we build its Python Wheel for you.

^{1. Our wheels for PyPy are exclusively available through pip and cannot be installed using conda.}

^{2. MacOS does not natively support NVIDIA GPUs.}

Supported GPU Architectures

imate can run on CUDA-capable multi-GPU devices. Using the docker container is the easiest way to run imate on GPU devices. The supported GPU micro-architectures and CUDA version are as follows:

Version \ Arch	Fermi	Kepler	Maxwell	Pascal	Volta	Turing	Ampere	Hopper
CUDA 9	✖	✖	✖	✖	✖	✖	✖	✖
CUDA 10	✖	✔	✔	✔	✔	✔	✔	✔
CUDA 11	✖	✖	✖	✔	✔	✔	✔	✔
CUDA 12	✖	✖	✖	✔	✔	✔	✔	✔

Documentation

See documentation, including:

Performance

imate is scalable to very large matrices. Its core library for basic linear algebraic operations is faster than OpenBLAS, and its pseudo-random generator is a hundred-fold faster than the implementation in the standard C++ library.

Read about the performance of imate in practical applications:

How to Contribute

We welcome contributions via GitHub’s pull request. If you do not feel comfortable modifying the code, we also welcome feature requests and bug reports as GitHub issues.

How to Cite

If you publish work that uses imate, please consider citing the manuscripts available here.

License

This project uses a BSD 3-clause license, in hopes that it will be accessible to most projects. If you require a different license, please raise an issue and we will consider a dual license.

Project details

These details have not been verified by PyPI

Project links

GitHub Statistics

View statistics for this project via Libraries.io, or by using our public dataset on Google BigQuery

Release history Release notifications | RSS feed

This version

0.25.2

Jan 23, 2024

0.25.1

Jan 23, 2024

0.25.0

Jan 22, 2024

0.23.4

Dec 21, 2023

0.23.1

Dec 20, 2023

0.23.0

Dec 20, 2023

0.22.1

Dec 18, 2023

0.22.0

Dec 17, 2023

0.21.5

Dec 15, 2023

0.21.4

Dec 15, 2023

0.21.3

Dec 14, 2023

0.21.1

Dec 14, 2023

0.21.0

Dec 13, 2023

0.20.14

Dec 4, 2023

0.20.11

Dec 3, 2023

0.20.10

Dec 1, 2023

0.20.9

Dec 1, 2023

0.20.6

Nov 30, 2023

0.20.5

Sep 20, 2023

0.20.4

Sep 19, 2023

0.20.3

Sep 18, 2023

0.20.1

Sep 16, 2023

0.20.0

Sep 14, 2023

0.19.3

Jan 12, 2023

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

imate-0.25.2.tar.gz (381.5 kB view hashes)

Uploaded Jan 23, 2024 Source

Built Distributions

imate-0.25.2-pp310-pypy310_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (3.8 MB view hashes)

Uploaded Jan 23, 2024 PyPy manylinux: glibc 2.17+ x86-64

imate-0.25.2-pp310-pypy310_pp73-macosx_11_0_arm64.whl (2.2 MB view hashes)

Uploaded Jan 23, 2024 PyPy macOS 11.0+ ARM64

imate-0.25.2-pp310-pypy310_pp73-macosx_10_9_x86_64.whl (2.4 MB view hashes)

Uploaded Jan 23, 2024 PyPy macOS 10.9+ x86-64

imate-0.25.2-pp39-pypy39_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (3.8 MB view hashes)

Uploaded Jan 23, 2024 PyPy manylinux: glibc 2.17+ x86-64

imate-0.25.2-pp39-pypy39_pp73-macosx_11_0_arm64.whl (2.2 MB view hashes)

Uploaded Jan 23, 2024 PyPy macOS 11.0+ ARM64

imate-0.25.2-pp39-pypy39_pp73-macosx_10_9_x86_64.whl (2.4 MB view hashes)

Uploaded Jan 23, 2024 PyPy macOS 10.9+ x86-64

imate-0.25.2-pp38-pypy38_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (3.8 MB view hashes)

Uploaded Jan 23, 2024 PyPy manylinux: glibc 2.17+ x86-64

imate-0.25.2-pp38-pypy38_pp73-macosx_10_9_x86_64.whl (2.4 MB view hashes)

Uploaded Jan 23, 2024 PyPy macOS 10.9+ x86-64

imate-0.25.2-cp312-cp312-win_amd64.whl (2.7 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.12 Windows x86-64

imate-0.25.2-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (4.0 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.12 manylinux: glibc 2.17+ x86-64

imate-0.25.2-cp312-cp312-macosx_11_0_arm64.whl (2.4 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.12 macOS 11.0+ ARM64

imate-0.25.2-cp312-cp312-macosx_10_9_x86_64.whl (2.6 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.12 macOS 10.9+ x86-64

imate-0.25.2-cp311-cp311-win_amd64.whl (2.7 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.11 Windows x86-64

imate-0.25.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (4.1 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.11 manylinux: glibc 2.17+ x86-64

imate-0.25.2-cp311-cp311-macosx_11_0_arm64.whl (2.4 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.11 macOS 11.0+ ARM64

imate-0.25.2-cp311-cp311-macosx_10_9_x86_64.whl (2.6 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.11 macOS 10.9+ x86-64

imate-0.25.2-cp310-cp310-win_amd64.whl (2.7 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.10 Windows x86-64

imate-0.25.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (4.1 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.10 manylinux: glibc 2.17+ x86-64

imate-0.25.2-cp310-cp310-macosx_11_0_arm64.whl (2.4 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.10 macOS 11.0+ ARM64

imate-0.25.2-cp310-cp310-macosx_10_9_x86_64.whl (2.7 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.10 macOS 10.9+ x86-64

imate-0.25.2-cp39-cp39-win_amd64.whl (2.8 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.9 Windows x86-64

imate-0.25.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (4.2 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.9 manylinux: glibc 2.17+ x86-64

imate-0.25.2-cp39-cp39-macosx_11_0_arm64.whl (2.4 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.9 macOS 11.0+ ARM64

imate-0.25.2-cp39-cp39-macosx_10_9_x86_64.whl (2.7 MB view hashes)

Uploaded Jan 23, 2024 CPython 3.9 macOS 10.9+ x86-64

Hashes for imate-0.25.2.tar.gz

Hashes for imate-0.25.2.tar.gz
Algorithm	Hash digest
SHA256	`3abce50f1e1a2a241db9dc98d259dd88480e38ac274fd44e31bac05634746aa9`
MD5	`11648b0032ff4cee205bba2c47d4579b`
BLAKE2b-256	`f228c23ca9643fb6d5f3b8b81a2139acda19a2c5c49290ae0845edfc885cacd4`

Hashes for imate-0.25.2-pp310-pypy310_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Hashes for imate-0.25.2-pp310-pypy310_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm	Hash digest
SHA256	`86bcc169ab06fa9f1841dcf15aa93a50de93af69838731b9811a1a96d8005592`
MD5	`475a4cda4193d916b2ec33f3ccfff985`
BLAKE2b-256	`b23be5042dafae722b21cb2e1b0ffb5e48bd6ca26a88bd93ea9cfbdb9b76eebc`

Hashes for imate-0.25.2-pp310-pypy310_pp73-macosx_11_0_arm64.whl

Hashes for imate-0.25.2-pp310-pypy310_pp73-macosx_11_0_arm64.whl
Algorithm	Hash digest
SHA256	`ff30d827ce1f6643e07fbcd1420869a39cf023cf85bb44ad5f465ee5ef9b9a2f`
MD5	`2343f9fa12de553a10234834392c1284`
BLAKE2b-256	`3ce3736250fee6a424a6690a26c0c8719ffa71ac97600d6b6d6f116f50dd3d7d`

Hashes for imate-0.25.2-pp310-pypy310_pp73-macosx_10_9_x86_64.whl

Hashes for imate-0.25.2-pp310-pypy310_pp73-macosx_10_9_x86_64.whl
Algorithm	Hash digest
SHA256	`2b01fa6ad942d81389e52f126ea326d77a638f03a6ae623a4636cdf579698f40`
MD5	`112feb0c7cbd5b6f306c5e8e136a2588`
BLAKE2b-256	`26ba8148e05cda020808efb1b3ee3513c7802d7098646410478e285cef43f997`

Hashes for imate-0.25.2-pp39-pypy39_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Hashes for imate-0.25.2-pp39-pypy39_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm	Hash digest
SHA256	`e35475c70935a59cda0aa0ef872c00f4e28e7e13e08562a760a10849c7ce98b6`
MD5	`43dca1a2fd784e30459ed606807fcb0a`
BLAKE2b-256	`ac5b97ec66b9c8e40f65e448a2178ed796c0a77488867afbca5aa8767921697d`

Hashes for imate-0.25.2-pp39-pypy39_pp73-macosx_11_0_arm64.whl

Hashes for imate-0.25.2-pp39-pypy39_pp73-macosx_11_0_arm64.whl
Algorithm	Hash digest
SHA256	`e924be8811e251511f628f7b77d531bac5e7147b13ce8d2e394a1e285c48f756`
MD5	`7b637bcf905af6b2f39231dc485ef004`
BLAKE2b-256	`f1b51697882eda09a9a13c4072194489b3eda0b99bbe8c0f181fcb146a71a452`

Hashes for imate-0.25.2-pp39-pypy39_pp73-macosx_10_9_x86_64.whl

Hashes for imate-0.25.2-pp39-pypy39_pp73-macosx_10_9_x86_64.whl
Algorithm	Hash digest
SHA256	`4df3c90ae796086da2070b0dd99e2e0bb3b6887cd6866799990f7c3d16880cde`
MD5	`43ddd1c8e908ef6adbd926f63609ef56`
BLAKE2b-256	`2798054d39cfa8ab511d986a36ad7e0e57447a4c92770f169e2f375e5eb20b32`

Hashes for imate-0.25.2-pp38-pypy38_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Hashes for imate-0.25.2-pp38-pypy38_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm	Hash digest
SHA256	`3be455d076b1f1199efb86013f72de97e9be9f5bdf5c08d82cea1c7d78f2f649`
MD5	`6940e92b7e0cbf6049d894a3a72d79d3`
BLAKE2b-256	`4777364fd7e43b3ad3101dfec5ea1d04e83a1dc8ee48ca4a36d7b3a4496da938`

Hashes for imate-0.25.2-pp38-pypy38_pp73-macosx_10_9_x86_64.whl

Hashes for imate-0.25.2-pp38-pypy38_pp73-macosx_10_9_x86_64.whl
Algorithm	Hash digest
SHA256	`f4678073abde16b9ae1f74a77a43d26d7d2c55ab94e12650240efc2957ae0075`
MD5	`59cf6429f711d707d7bac563dca53208`
BLAKE2b-256	`5dc2d40c9cd605df655fe03ef9101c19b74d588d4906f95b3ad0c41ee166ffcf`

Hashes for imate-0.25.2-cp312-cp312-win_amd64.whl

Hashes for imate-0.25.2-cp312-cp312-win_amd64.whl
Algorithm	Hash digest
SHA256	`db4c859b7aab0dc115ed976789485137c48fb0c0ff2737579d0dd2626c559ab3`
MD5	`3183a5a62d5f1d73824258cb456b46e3`
BLAKE2b-256	`bfe32a6ad932e9700c9898c70b653ac4cc18785064e45a1956b7e8d12fff8aee`

Hashes for imate-0.25.2-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Hashes for imate-0.25.2-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm	Hash digest
SHA256	`c37ac5fc298fea06d714d3ad0591ef6c43c65843f26250a034b87d86f80f6d84`
MD5	`b47df917045fe820a7580d007f009fc6`
BLAKE2b-256	`6cc9dfa03dafa4900e2b294f2fe90ea602f770123d0fce57eb64338fdf298660`

Hashes for imate-0.25.2-cp312-cp312-macosx_11_0_arm64.whl

Hashes for imate-0.25.2-cp312-cp312-macosx_11_0_arm64.whl
Algorithm	Hash digest
SHA256	`7129f8348cf90cf38a64289aa1043fa9993d6d57c68cd7c2c62dc82f7e405376`
MD5	`7e17cd2bc7267ecfd09171645dfe8cf9`
BLAKE2b-256	`c4b0be7236e52bf02f77a882ae74481e0603c309898d2bb3f2118ec8330f63e1`

Hashes for imate-0.25.2-cp312-cp312-macosx_10_9_x86_64.whl

Hashes for imate-0.25.2-cp312-cp312-macosx_10_9_x86_64.whl
Algorithm	Hash digest
SHA256	`87fcd35a9c3d4688af4a04c29c2a4029d801da154dcfe688635606c678529483`
MD5	`02da3f61638a51a97cc71143346a5a7d`
BLAKE2b-256	`c6f6b90d9c904c2ad3d92aa6c7e431328a50466100672b7ba289fe899ce68f76`

Hashes for imate-0.25.2-cp311-cp311-win_amd64.whl

Hashes for imate-0.25.2-cp311-cp311-win_amd64.whl
Algorithm	Hash digest
SHA256	`082b7f1a19f7ec0a024c9efa9f7a4e05601aae9ded5293e134a7e768ae5b8a5d`
MD5	`bc3187546f6448015c6c4f07d4694f10`
BLAKE2b-256	`1a75c5dde2302222d151ac499934c3db2d96d7e587b196eb2f32bb00f5c87606`

Hashes for imate-0.25.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Hashes for imate-0.25.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm	Hash digest
SHA256	`995d809253ea57ee55f99f83306cc1789484d75702658bb85827cfe2153c86b3`
MD5	`fcfa3e1b2db3c6562d74a5e920acd054`
BLAKE2b-256	`e398f72b9bebf6c7bde29c1ece1ca7f30b50f21fbb54fbaaed21c1cc2a998d85`

Hashes for imate-0.25.2-cp311-cp311-macosx_11_0_arm64.whl

Hashes for imate-0.25.2-cp311-cp311-macosx_11_0_arm64.whl
Algorithm	Hash digest
SHA256	`f25c5eb9a6666520ceb0c9ab54436c020ed6da66b276b6a0504ef08a6400dc6d`
MD5	`2788cc32e031e2e25f454cc50f7975a5`
BLAKE2b-256	`abd7476ba3e9fc9a6fd2b91ec9f053adcc8d49057e26fdde955a46dfc02cbb82`

Hashes for imate-0.25.2-cp311-cp311-macosx_10_9_x86_64.whl

Hashes for imate-0.25.2-cp311-cp311-macosx_10_9_x86_64.whl
Algorithm	Hash digest
SHA256	`99fc19ced3e89721b2f08cdcfdc7fcf7164c713ac7686fce49de227a14a31e91`
MD5	`44fb212594e4a6708321ea58104b4708`
BLAKE2b-256	`dd3832e84ef8f922ffc1c926c6417cd638913273851362fc37118b56025c3339`

Hashes for imate-0.25.2-cp310-cp310-win_amd64.whl

Hashes for imate-0.25.2-cp310-cp310-win_amd64.whl
Algorithm	Hash digest
SHA256	`9318156142aa8efedb1e032f86e2f00dc170566e9c099ec56397b919b5d9efe8`
MD5	`c6b28e3944e0064bd493876d48e25370`
BLAKE2b-256	`37977d1e965d6c5166d96d0a83ce8713c2172bf3e8eea166e4deb6c229a3f05a`

Hashes for imate-0.25.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Hashes for imate-0.25.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm	Hash digest
SHA256	`ec836c2cfe497de1f0eeca7e4dac59c6dd4c89dec510e535ebe43b613ac81725`
MD5	`6c7f9056573b155d62db9fecdda16685`
BLAKE2b-256	`e19f14790433eb0e25441b6ee708bad165614945f36c2e8cb66f7d46d8b2e50d`

Hashes for imate-0.25.2-cp310-cp310-macosx_11_0_arm64.whl

Hashes for imate-0.25.2-cp310-cp310-macosx_11_0_arm64.whl
Algorithm	Hash digest
SHA256	`14852cdf0ee8ce8611bf6af2997cafd7d5c9a13256c2ce0b6f77008d6a1e884d`
MD5	`2c02e46914c0eb971ef2743ba453b922`
BLAKE2b-256	`864a08ab3572bab0088cc47f9178beef4a5c5c8648bd2b503c74e28d69be2fd4`

Hashes for imate-0.25.2-cp310-cp310-macosx_10_9_x86_64.whl

Hashes for imate-0.25.2-cp310-cp310-macosx_10_9_x86_64.whl
Algorithm	Hash digest
SHA256	`05c1bfe6e5ce0850d4e271a50f931860f3148783269e0e101e376866ddf87ad1`
MD5	`0a108fac273a1375a5340d436b1bb08e`
BLAKE2b-256	`d41e32f825d96336c6932b3f702624d3f66040c13d26537f1b57ac5ae18772fb`

Hashes for imate-0.25.2-cp39-cp39-win_amd64.whl

Hashes for imate-0.25.2-cp39-cp39-win_amd64.whl
Algorithm	Hash digest
SHA256	`10710e4ee90e5c77118a00e9f6d765f4f7e2c454c908b432f35272d9f7a8a76c`
MD5	`831f1d56312494160f3db488c93286aa`
BLAKE2b-256	`271c6f1b4cb9b367d4bebdd10e2b88f7367b3f6bb586f4101426940e50462d33`

Hashes for imate-0.25.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Hashes for imate-0.25.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm	Hash digest
SHA256	`87d877c19f8e78891c7f4d0e5ef56b454c4eaeece3c32735d90cda663f7227dc`
MD5	`f6402820490463d4a5553fb197c43dc3`
BLAKE2b-256	`82095d53f3c2f98f94da79c6f33a15f45b1a824188e9726dbef9f4a924de2159`

Hashes for imate-0.25.2-cp39-cp39-macosx_11_0_arm64.whl

Hashes for imate-0.25.2-cp39-cp39-macosx_11_0_arm64.whl
Algorithm	Hash digest
SHA256	`1303f6926fa76c0cd1bd4ba2b31f975db922987a788357663265e813d18990a1`
MD5	`ccbe7481e3f07eaec85117b78369b39f`
BLAKE2b-256	`c9d88f49532547e13ab92676d1cacc853d38cda3bc9738152307c739b90e4c84`

Hashes for imate-0.25.2-cp39-cp39-macosx_10_9_x86_64.whl

Hashes for imate-0.25.2-cp39-cp39-macosx_10_9_x86_64.whl
Algorithm	Hash digest
SHA256	`e6baa78b8adf9c64aff1783061edb5501f01f9d386e6fb7fe0f02a2062d89680`
MD5	`f3f15a4c558989c0b8eda531a46c3bf7`
BLAKE2b-256	`309d3fda67401a6ea2c7237e12774e97b0707dbcb14560b514f0e3d8ec7986a8`