PyLops-MPI - MPI Powered PyLops with mpi4py
PyLops-MPI - MPI Powered PyLops with mpi4py - Published in JOSS (2025)
Science Score: 95.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
✓DOI references
Found 4 DOI reference(s) in README and JOSS metadata -
✓Academic publication links
Links to: joss.theoj.org -
✓Committers with academic emails
1 of 5 committers (20.0%) from academic institutions -
○Institutional organization owner
-
✓JOSS paper metadata
Published in Journal of Open Source Software
Repository
MPI-powered PyLops with MPI4Py
Basic Info
- Host: GitHub
- Owner: PyLops
- License: lgpl-3.0
- Language: Python
- Default Branch: main
- Homepage: https://pylops.github.io/pylops-mpi/
- Size: 101 MB
Statistics
- Stars: 22
- Watchers: 1
- Forks: 5
- Open Issues: 7
- Releases: 5
Metadata Files
README.md
Distributed linear operators and solvers
Pylops-mpi is a Python library built on top of PyLops, designed to enable distributed and parallel processing of large-scale linear algebra operations and computations.
Installation
To install pylops-mpi, you need to have Message Passing Interface (MPI) and optionally Nvidia's Collective Communication Library (NCCL) installed on your system.
Download and Install MPI: Visit the official MPI website to download an appropriate MPI implementation for your system. Follow the installation instructions provided by the MPI vendor.
Verify MPI Installation: After installing MPI, verify its installation by opening a terminal or command prompt and running the following command:
mpiexec --versionInstall pylops-mpi: Once MPI is installed and verified, you can proceed to install
pylops-mpiviapip:pip install pylops-mpi(Optional) To enable the NCCL backend for multi-GPU systems, install
cupyandncclviapip:pip install cupy-cudaXx nvidia-nccl-cuX
with X=11,12.
Alternatively, if the Conda package manager is used to setup the Python environment, steps 1 and 2 can be skipped and mpi4py can be installed directly alongside the MPI distribution of choice:
conda install -c conda-forge mpi4py X
with X=mpich, openmpi, impi_rt, msmpi. Similarly step 4 can be accomplished using:
conda install -c conda-forge cupy nccl
See the docs (Installation) for more information.
Run Pylops-MPI
Once you have installed the prerequisites and pylops-mpi, you can run pylops-mpi using the mpiexec command.
Here is an example on how to run a python script called <script_name>.py:
mpiexec -n <NUM_PROCESSES> python <script_name>.py
Example: A distributed finite-difference operator
The following example is a modified version of
PyLops' README_ starting
example that can handle a 2D-array distributed across ranks over the first dimension
via the DistributedArray object:
```python import numpy as np from pylops_mpi import DistributedArray, Partition
Initialize DistributedArray with partition set to Scatter
nx, ny = 11, 21 x = np.zeros((nx, ny), dtype=np.float64) x[nx // 2, ny // 2] = 1.0
xdist = pylopsmpi.DistributedArray.to_dist( x=x.flatten(), partition=Partition.SCATTER)
Distributed first-derivative
Dop = pylopsmpi.MPIFirstDerivative((nx, ny), dtype=np.float64)
y = Dx
ydist = Dop @ x_dist
xadj = D^H y
xadjdist = Dop.H @ y_dist
xinv = D^-1 y
x0dist = pylopsmpi.DistributedArray(Dop.shape[1], dtype=np.float64) x0dist[:] = 0 xinvdist = pylopsmpi.cgls(Dop, ydist, x0=x0_dist, niter=10)[0] ```
Note that the DistributedArray class provides the to_dist class method that accepts a NumPy array as input and converts it into an instance of the DistributedArray class. This method is used to transform a regular NumPy array into a DistributedArray that is distributed and processed across multiple nodes or processes.
Moreover, the DistributedArray class provides also fundamental mathematical operations, such as element-wise addition, subtraction, multiplication, dot product, and an equivalent of the np.linalg.norm function that operate in a distributed fashion,
thus utilizing the efficiency of the MPI/NCC; protocols. This enables efficient computation and processing of large-scale distributed arrays.
Running Tests
The MPI test scripts are located in the tests folder.
Use the following command to run the tests:
mpiexec -n <NUM_PROCESSES> pytest tests/ --with-mpi
where the --with-mpi option tells pytest to enable the pytest-mpi plugin, allowing the tests to utilize the MPI functionality.
Similarly, to run the NCCL test scripts in the tests_nccl folder,
use the following command to run the tests:
mpiexec -n <NUM_PROCESSES> pytest tests_nccl/ --with-mpi
Documentation
The official documentation of Pylops-MPI is available here. Visit the official docs to learn more about pylops-mpi.
Contributors
- Rohan Babbar, rohanbabbar04
- Yuxi Hong, hongyx11
- Matteo Ravasi, mrava87
- Tharit Tangkijwanichakul, tharittk
Owner
- Name: PyLops
- Login: PyLops
- Kind: organization
- Repositories: 14
- Profile: https://github.com/PyLops
Matrix-Free linear algebra and optimization in Python
JOSS Publication
PyLops-MPI - MPI Powered PyLops with mpi4py
Authors
Cluster Innovation Center, University of Delhi, Delhi, India.
Earth Science and Engineering, Physical Sciences and Engineering (PSE), King Abdullah University of Science and Technology (KAUST), Thuwal, Kingdom of Saudi Arabia.
Lawrence Berkeley National Laboratory, Berkeley, California, United States of America.
Editor
George K. Thiruvathukal
Tags
MPI High Performance ComputingGitHub Events
Total
- Create event: 18
- Release event: 3
- Issues event: 15
- Watch event: 8
- Delete event: 15
- Member event: 1
- Issue comment event: 74
- Push event: 89
- Pull request review comment event: 199
- Pull request review event: 159
- Pull request event: 64
- Fork event: 4
Last Year
- Create event: 18
- Release event: 3
- Issues event: 15
- Watch event: 8
- Delete event: 15
- Member event: 1
- Issue comment event: 74
- Push event: 89
- Pull request review comment event: 199
- Pull request review event: 159
- Pull request event: 64
- Fork event: 4
Committers
Last synced: 7 months ago
Top Committers
| Name | Commits | |
|---|---|---|
| rohanbabbar04 | r****8@y****m | 207 |
| mrava87 | m****i@g****m | 92 |
| tharittk | t****j@g****m | 62 |
| astroC86 | 6****6 | 22 |
| Tharit Tangkijwanichakul | t****l@d****u | 1 |
Committer Domains (Top 20 + Academic)
Issues and Pull Requests
Last synced: 6 months ago
All Time
- Total issues: 31
- Total pull requests: 150
- Average time to close issues: 19 days
- Average time to close pull requests: 4 days
- Total issue authors: 3
- Total pull request authors: 5
- Average comments per issue: 1.65
- Average comments per pull request: 1.63
- Merged pull requests: 123
- Bot issues: 0
- Bot pull requests: 0
Past Year
- Issues: 10
- Pull requests: 86
- Average time to close issues: 14 days
- Average time to close pull requests: 4 days
- Issue authors: 3
- Pull request authors: 5
- Average comments per issue: 1.9
- Average comments per pull request: 1.06
- Merged pull requests: 61
- Bot issues: 0
- Bot pull requests: 0
Top Authors
Issue Authors
- mrava87 (25)
- rohanbabbar04 (5)
- tharittk (1)
Pull Request Authors
- mrava87 (62)
- rohanbabbar04 (60)
- tharittk (20)
- astroC86 (6)
- danielskatz (2)
Top Labels
Issue Labels
Pull Request Labels
Packages
- Total packages: 1
-
Total downloads:
- pypi 168 last-month
- Total dependent packages: 0
- Total dependent repositories: 0
- Total versions: 5
- Total maintainers: 1
pypi.org: pylops-mpi
Python library implementing linear operators with MPI
- Documentation: https://pylops-mpi.readthedocs.io/
- License: GNU LESSER GENERAL PUBLIC LICENSE Version 3, 29 June 2007 Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/> Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed. This version of the GNU Lesser General Public License incorporates the terms and conditions of version 3 of the GNU General Public License, supplemented by the additional permissions listed below. 0. Additional Definitions. As used herein, "this License" refers to version 3 of the GNU Lesser General Public License, and the "GNU GPL" refers to version 3 of the GNU General Public License. "The Library" refers to a covered work governed by this License, other than an Application or a Combined Work as defined below. An "Application" is any work that makes use of an interface provided by the Library, but which is not otherwise based on the Library. Defining a subclass of a class defined by the Library is deemed a mode of using an interface provided by the Library. A "Combined Work" is a work produced by combining or linking an Application with the Library. The particular version of the Library with which the Combined Work was made is also called the "Linked Version". The "Minimal Corresponding Source" for a Combined Work means the Corresponding Source for the Combined Work, excluding any source code for portions of the Combined Work that, considered in isolation, are based on the Application, and not on the Linked Version. The "Corresponding Application Code" for a Combined Work means the object code and/or source code for the Application, including any data and utility programs needed for reproducing the Combined Work from the Application, but excluding the System Libraries of the Combined Work. 1. Exception to Section 3 of the GNU GPL. You may convey a covered work under sections 3 and 4 of this License without being bound by section 3 of the GNU GPL. 2. Conveying Modified Versions. If you modify a copy of the Library, and, in your modifications, a facility refers to a function or data to be supplied by an Application that uses the facility (other than as an argument passed when the facility is invoked), then you may convey a copy of the modified version: a) under this License, provided that you make a good faith effort to ensure that, in the event an Application does not supply the function or data, the facility still operates, and performs whatever part of its purpose remains meaningful, or b) under the GNU GPL, with none of the additional permissions of this License applicable to that copy. 3. Object Code Incorporating Material from Library Header Files. The object code form of an Application may incorporate material from a header file that is part of the Library. You may convey such object code under terms of your choice, provided that, if the incorporated material is not limited to numerical parameters, data structure layouts and accessors, or small macros, inline functions and templates (ten or fewer lines in length), you do both of the following: a) Give prominent notice with each copy of the object code that the Library is used in it and that the Library and its use are covered by this License. b) Accompany the object code with a copy of the GNU GPL and this license document. 4. Combined Works. You may convey a Combined Work under terms of your choice that, taken together, effectively do not restrict modification of the portions of the Library contained in the Combined Work and reverse engineering for debugging such modifications, if you also do each of the following: a) Give prominent notice with each copy of the Combined Work that the Library is used in it and that the Library and its use are covered by this License. b) Accompany the Combined Work with a copy of the GNU GPL and this license document. c) For a Combined Work that displays copyright notices during execution, include the copyright notice for the Library among these notices, as well as a reference directing the user to the copies of the GNU GPL and this license document. d) Do one of the following: 0) Convey the Minimal Corresponding Source under the terms of this License, and the Corresponding Application Code in a form suitable for, and under terms that permit, the user to recombine or relink the Application with a modified version of the Linked Version to produce a modified Combined Work, in the manner specified by section 6 of the GNU GPL for conveying Corresponding Source. 1) Use a suitable shared library mechanism for linking with the Library. A suitable mechanism is one that (a) uses at run time a copy of the Library already present on the user's computer system, and (b) will operate properly with a modified version of the Library that is interface-compatible with the Linked Version. e) Provide Installation Information, but only if you would otherwise be required to provide such information under section 6 of the GNU GPL, and only to the extent that such information is necessary to install and execute a modified version of the Combined Work produced by recombining or relinking the Application with a modified version of the Linked Version. (If you use option 4d0, the Installation Information must accompany the Minimal Corresponding Source and Corresponding Application Code. If you use option 4d1, you must provide the Installation Information in the manner specified by section 6 of the GNU GPL for conveying Corresponding Source.) 5. Combined Libraries. You may place library facilities that are a work based on the Library side by side in a single library together with other library facilities that are not Applications and are not covered by this License, and convey such a combined library under terms of your choice, if you do both of the following: a) Accompany the combined library with a copy of the same work based on the Library, uncombined with any other library facilities, conveyed under the terms of this License. b) Give prominent notice with the combined library that part of it is a work based on the Library, and explaining where to find the accompanying uncombined form of the same work. 6. Revised Versions of the GNU Lesser General Public License. The Free Software Foundation may publish revised and/or new versions of the GNU Lesser General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. Each version is given a distinguishing version number. If the Library as you received it specifies that a certain numbered version of the GNU Lesser General Public License "or any later version" applies to it, you have the option of following the terms and conditions either of that published version or of any later version published by the Free Software Foundation. If the Library as you received it does not specify a version number of the GNU Lesser General Public License, you may choose any version of the GNU Lesser General Public License ever published by the Free Software Foundation. If the Library as you received it specifies that a proxy can decide whether future versions of the GNU Lesser General Public License shall apply, that proxy's public statement of acceptance of any version is permanent authorization for you to choose that version for the Library.
-
Latest release: 0.3.0
published 7 months ago