An error occurred while retrieving target projects.

Compare revisions

Changes are shown as if the source revision was being merged into the target revision. Learn more about comparing revisions.

Source

Target

Show changes

Only incoming changes from source

Include changes to target since source was created

Commits on Source 762

637684c1 Fix OpenCL for CUDA OpenCL by avoid int(...) constructor
Dec 5, 2019
d3fce803 Merge branch 'fix-llvm-gpu' into 'master'
Dec 5, 2019
3d9e0c6d Merge branch 'fix-opencl' into 'master'
Dec 5, 2019
efadec2e Minor improvements to FiniteDifferenceStaggeredStencilDerivation
Dec 6, 2019
0684cfe3 Merge branch 'fvm' into 'master'
Dec 6, 2019
ed864c0e Field.Access consistently store offsets in sympyfied form
Dec 6, 2019
5790e28d Fix FiniteDifferenceStaggeredStencilDerivation on Sympy 1.1
Dec 6, 2019
66467409 Merge branch 'fvm' into 'master'
Dec 6, 2019
f9ba7391 Fixed tutorial notebook 06
Dec 10, 2019
7d77f19b Allow `CUSTOM_FIELD`s to have a different size in transformations.py
Dec 16, 2019
212776a3 Implement __bool__ for AssignmentCollection
Dec 17, 2019
fbf1ec12 Assert non-empty assignments on kernelcreation
Dec 17, 2019
01cdfa64 Merge branch 'custom-fields-different-size-in-transformations' into 'master'
Dec 17, 2019
00c00c45 Merge branch 'AssignmentCollection.__bool__' into 'master'
Dec 17, 2019
37aff58b FiniteDifferenceStaggeredStencilDerivation must be applied to field access
Dec 17, 2019
4751e6cb Update AssignmentCollection.__repr__
Dec 17, 2019
1e02ab8f fix staggered field access prettyprinter
Dec 17, 2019
29ad4e74 Merge branch 'update-AssignmentCollection.__repr__' into 'master'
Dec 18, 2019
f9e88655 Merge branch 'fvm' into 'master'
Dec 18, 2019
ef543c5e Fix AssignmentCollection.{free_symbols,bound_symbols,defined_symbols} for non-Assignments
Dec 19, 2019
aa6dfe57 Add a star to correctly get atoms of InterpolatorAccess
Dec 19, 2019
87913d0c Merge branch 'star-star-star' into 'master'
Dec 19, 2019
e512cf88 Merge branch 'fix-free/defined-symbols-for-non-Assignments' into 'master'
Dec 19, 2019
8aa4fe9f Allow functions for Field.coordinate_transform
Jan 3, 2020
5257b4ae Also update Field.physical_to_index,index_to_physical for coordinate_transform as a function
Jan 8, 2020
5d3e1948 Fixed bug in sympy intersphinx
Jan 9, 2020
9b57cf87 Merge branch 'functional_coordinate_transform' into 'master'
Jan 9, 2020
de753874 Improve OpenCL integration via create_kernel
Jan 10, 2020
e6a30483 Extend SerialDataHandling to handle alternative GPU libraries
Jan 10, 2020
061ab58b Allow creation of indexed kernels with opencl
Jan 10, 2020
e02acf7d Let SerialDataHandling.synchronization_function work with target OpenCL
Jan 10, 2020
585084e7 Teach SympyPrinters how to print Type
Jan 10, 2020
0e56c1e3 Make SerialDataHandling.serialization_function work with OpenCL
Jan 10, 2020
0b0ba401 Add two helper functions to manage global cl ctx
Jan 10, 2020
5437459f Add pystencils.opencl.autoinit
Jan 10, 2020
2901a5b5 Protect pystencils.opencl.autoinit with try..except
Jan 10, 2020
62803d18 Add pycuda.autoinit to GPU test
Jan 10, 2020
de710402 Allow default_target=='opencl' in SerialDataHandling
Jan 10, 2020
58e62e71 Add friendly hint when calling OpenCL kernels with PyCUDA/Numpy arrays
Jan 10, 2020
bb21c9b1 Add weird construction to handle OpenCL boundary handling (don't show
Jan 10, 2020
d08987c1 Conftest ignore opencl.autoinit
Jan 10, 2020
c989d8ae Simplify boundary handling with OpenCL
Jan 10, 2020
226abc3a test_staggered_kernel also for OpenCL
Jan 10, 2020
171d844a Merge branch 'opencl-datahandling' into 'master'
Jan 10, 2020
1b4ace3e Add DataHandling.add_arrays
Jan 13, 2020
e5f4c74d Make add_arrays return fields to arrays
Jan 13, 2020
023e9659 Add type annotation for DataHandling.add_arrays
Jan 14, 2020
1af235da Automatic Finite Volume and Volume-of-Fluid discretization
Jan 14, 2020
5d0a131c Merge branch 'fvm' into 'master'
Jan 14, 2020
b6c9f64c Merge branch 'Datahandling.add_arrays' into 'master'
Jan 14, 2020
47dbf5d3 SerialDataHandling: synchronization_function for tensor fields
Jan 14, 2020
4b618138 Make create_staggered_kernel work with OpenMP
Jan 14, 2020
99d4b9fc Merge branch 'staggered' into 'master'
Jan 14, 2020
245e8331 Add dtype to DataHandling.add_arrays
Jan 15, 2020
d0064ffb Merge branch 'dtype-for-add_arrays' into 'master'
Jan 15, 2020
4dcc3ea5 Fix {PyCuda,PyOpenCl}ArrayHandler.{zeros,ones,randn}
Jan 15, 2020
d49f555a Merge branch 'fix-arrayhandlers' into 'master'
Jan 15, 2020
cf580200 Fix failing interpolation tests on newer sympy versions
Jan 16, 2020
f278b444 Fix in operation counting - FLOPs inside division were not counted
Jan 16, 2020
91cae896 Added option to pass options to sympy CSE through simplification rule
Jan 16, 2020
fc81c5ae Merge branch 'sync' into 'master'
Jan 16, 2020
8cca242f Merge branch 'interpolation-refactoring' into 'master'
Jan 16, 2020
b63cb073 Fix for lbmpy sparse setups - there kernel with only custom fields exist
Jan 16, 2020
659319bd Correctly print sp.Abs in generated code
Jan 16, 2020
050107e0 Bugfix: pystencils.Type is a sympy.Atom now
Jan 16, 2020
f4abda02 Fix in boundary handling - check for GPU_LIKE_TARGETS
Jan 16, 2020
b0aae533 Boundary Handling fix to work with ParallelDatahandling again
Jan 16, 2020
49e5c430 Fix FiniteDifferenceStencilDerivation.rotate_weights_and_apply
Jan 16, 2020
59e3861e Fix bug related to ParallelDataHandling
Jan 17, 2020
f4008b32 Print actual block contents in Conditional.{__repr__,__str__}
Jan 17, 2020
f5db4b53 Merge branch 'Conditional.__repr__' into 'master'
Jan 17, 2020
5d569c84 Merge branch 'fix-rotate_weights_and_apply' into 'master'
Jan 17, 2020
f16225d5 Merge branch 'fix-ParallelDataHandling-bug!' into 'master'
Jan 17, 2020
1e7a179a Fix #6: let show_code display/print code
Jan 18, 2020
38da1c39 Fix #17 : Allow vector assignments
Jan 18, 2020
1c57a059 Assert same length when performing vector assignment
Jan 18, 2020
8568124d cuda.TextureReference.set_array(tex_ref, cu_array) -> tex_ref.set_array(cu_array)
Jan 19, 2020
bec36bb2 Rename: get_code -> get_code_obj
Jan 20, 2020
3078eb30 CI: Replace minimal-ubuntu job with ubuntu
Jan 20, 2020
35db19ff Add typeannotations and docstring for KernelWrapper
Jan 21, 2020
2e7f2703 Auto-format conftest
Jan 21, 2020
bbb7ab95 Merge branch 'kernel-wrapper-typeannotations' into 'master'
Jan 21, 2020
75b3dad8 Merge branch 'show_code' into 'master'
Jan 21, 2020
46b7b88d Merge branch 'vector-assignments' into 'master'
Jan 21, 2020
01a56d31 skip tests with unmet dependencies
Jan 22, 2020
607aafa8 fix more tests on Ubuntu
Jan 22, 2020
2c65b1f8 Remove clear_global_ctx in OpenCL test (apparently createing two CL contexts does not work always)
Jan 22, 2020
7cc8270b Remove import of pytest in test_datahandling (no pytest installed on
Jan 22, 2020
be937958 Cherry-pick changes from interpolation-refactoring
Jan 22, 2020
fe7b4cd0 xfail failing gpu interpolation tests
Jan 22, 2020
85f9aa58 Require cuda for ubuntu CI. Since it is used
Jan 22, 2020
579bc181 Probably also AVX
Jan 22, 2020
0b15e10c Donot save image in interpolation tests
Jan 23, 2020
d69f5d11 xfail and skip (due to old pytest version)
Jan 23, 2020
2872b6f0 Importorskip in test_interpolation
Jan 23, 2020
beb22654 correct type conversion in generated complex code
Jan 23, 2020
da296e88 Merge branch 'oldsympy' of i10git.cs.fau.de:pycodegen/pystencils into oldsympy
Jan 23, 2020
4bae07a3 flake8 complex numbers
Jan 23, 2020
f6f19153 correctly skip interpolation test without CUDA
Jan 23, 2020
94caee94 skip more tests based on dependencies
Jan 23, 2020

662 additional commits have been omitted to prevent performance issues.

Some changes are not shown.

For a faster browsing experience, only 56 of 288 files are shown. Download one of the files below to see all changes.

.flake8

+4 −4

Original line number	Diff line number	Diff line
		[flake8]
		max-line-length=120
		exclude=pystencils/jupyter.py,
		pystencils/plot.py
		pystencils/session.py
		ignore = W293 W503 W291 C901
		exclude=src/pystencils/jupyter.py,
		src/pystencils/plot.py
		src/pystencils/session.py
		ignore = W293 W503 W291 C901 E741

.gitattributes

0 → 100644

+1 −0

Original line number	Diff line number	Diff line
		src/pystencils/_version.py export-subst

.gitignore

+17 −2

Original line number	Diff line number	Diff line
		__pycache__
		.ipynb_checkpoints
		.coverage
		.coverage*
		*.pyc
		*.vti
		/build
		/dist
		/*.egg-info
		*.egg-info
		.cache
		_build
		/html_doc
		/.idea
		.vscode
		.cache
		_local_tmp
		RELEASE-VERSION
		test-report
		src/pystencils/boundaries/createindexlistcython.c
		src/pystencils/boundaries/createindexlistcython.*.so
		tests/tmp
		tests/var
		tests/kerncraft_inputs/.2d-5pt.c_kerncraft/
		tests/kerncraft_inputs/.3d-7pt.c_kerncraft/
		report.xml
		coverage_report/


		# macOS
		**/.DS_Store
		*.uuid

.gitlab-ci.yml

+241 −63

Original line number	Diff line number	Diff line
		stages:
		- pretest
		- test
		- nightly
		- docs
		- deploy


		# -------------------------- Templates ------------------------------------------------------------------------------------

		# Base configuration for jobs meant to run at every commit
		.every-commit:
		rules:
		- if: $CI_PIPELINE_SOURCE != "schedule"

		# Configuration for jobs meant to run on each commit to pycodegen/pystencils/master
		.every-commit-master:
		rules:
		- if: '$CI_PIPELINE_SOURCE != "schedule" && $CI_PROJECT_PATH == "pycodegen/pystencils" && $CI_COMMIT_BRANCH == "master"'

		# Base configuration for jobs meant to run at a schedule
		.scheduled:
		rules:
		- if: $CI_PIPELINE_SOURCE == "schedule"

		# -------------------------- Tests ------------------------------------------------------------------------------------

		# Normal test - runs on every commit all but "long run" tests
		tests-and-coverage:
		stage: test
		except:
		variables:
		- $ENABLE_NIGHTLY_BUILDS
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full
		stage: pretest
		extends: .every-commit
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full:cupy12.3
		before_script:
		- pip install -e .
		script:
		- env
		- pip list
		- export NUM_CORES=$(nproc --all)
		- mkdir -p ~/.config/matplotlib
		- echo "backend:template" > ~/.config/matplotlib/matplotlibrc
		- mkdir public
		- py.test -v -n $NUM_CORES --cov-report html --cov-report term --cov=. -m "not longrun" --html test-report/index.html
		- pytest -v -n $NUM_CORES --cov-report html --cov-report xml --cov-report term --cov=. -m "not longrun" --html test-report/index.html --junitxml=report.xml
		- python -m coverage xml
		tags:
		- docker
		- cuda
		- cuda11
		- AVX
		coverage: /Total coverage:\s\d+.\d+\%/
		artifacts:
		when: always
		paths:
		- coverage_report
		- test-report
		reports:
		coverage_report:
		coverage_format: cobertura
		path: coverage.xml
		junit: report.xml

		# Nightly test - runs "long run" jobs only
		test-longrun:
		# Normal test with longruns
		tests-and-coverage-with-longrun:
		stage: test
		only:
		variables:
		- $ENABLE_NIGHTLY_BUILDS
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full
		when: manual
		allow_failure: true
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full:cupy12.3
		before_script:
		- pip install sympy --upgrade
		- pip install -e .
		script:
		- env
		- pip list
		- export NUM_CORES=$(nproc --all)
		- mkdir -p ~/.config/matplotlib
		- echo "backend:template" > ~/.config/matplotlib/matplotlibrc
		- py.test -v -n $NUM_CORES --cov-report html --cov-report term --cov=. --html test-report/index.html
		- mkdir public
		- py.test -v -n $NUM_CORES
		tags:
		- docker
		- cuda11
		- AVX

		# pipeline with latest python version
		latest-python:
		stage: test
		extends: .every-commit
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/latest_python
		before_script:
		- pip install -e .
		script:
		- env
		- pip list
		- pip install -e .
		- export NUM_CORES=$(nproc --all)
		- mkdir -p ~/.config/matplotlib
		- echo "backend:template" > ~/.config/matplotlib/matplotlibrc
		- mkdir public
		- py.test -v -n $NUM_CORES -m "not longrun" --junitxml=report.xml
		tags:
		- docker
		- cuda
		- AVX
		artifacts:
		paths:
		- coverage_report
		- test-report
		when: always
		reports:
		junit: report.xml


		# Minimal tests in windows environment
		minimal-windows:
		#minimal-windows:
		# stage: test
		# tags:
		# - win
		# script:
		# - export NUM_CORES=$(nproc --all)
		# - source /cygdrive/c/Users/build/Miniconda3/Scripts/activate
		# - source activate pystencils
		# - pip install joblib
		# - pip list
		# - python -c "import numpy"
		# - py.test -v -m "not (notebook or longrun)"

		ubuntu:
		stage: test
		except:
		variables:
		- $ENABLE_NIGHTLY_BUILDS
		tags:
		- win
		extends: .every-commit
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/ubuntu
		before_script:
		- ln -s /usr/include/locale.h /usr/include/xlocale.h
		- pip3 install -e .
		script:
		- source /cygdrive/c/Users/build/Miniconda3/Scripts/activate
		- source activate pystencils_dev
		- export NUM_CORES=$(nproc --all)
		- mkdir -p ~/.config/matplotlib
		- echo "backend:template" > ~/.config/matplotlib/matplotlibrc
		- sed -i 's/--doctest-modules //g' pytest.ini
		- env
		- conda env list
		- python -c "import numpy"
		- python setup.py quicktest
		- pip list
		- pytest -v -n $NUM_CORES -m "not longrun" --junitxml=report.xml
		tags:
		- docker
		- cuda11
		- AVX
		artifacts:
		when: always
		reports:
		junit: report.xml

		minimal-ubuntu:
		.multiarch_template:
		stage: test
		except:
		variables:
		- $ENABLE_NIGHTLY_BUILDS
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/minimal_ubuntu
		extends: .every-commit
		allow_failure: true
		before_script: &multiarch_before_script
		# - pip3 install -v .
		- export PYTHONPATH=src
		- python3 -c "import pystencils as ps; ps.cpu.cpujit.read_config()"
		- sed -i '/^fail_under.*/d' pytest.ini
		script:
		- python3 setup.py quicktest
		- export NUM_CORES=$(nproc --all)
		- mkdir -p ~/.config/matplotlib
		- echo "backend:template" > ~/.config/matplotlib/matplotlibrc
		- sed -i 's/--doctest-modules //g' pytest.ini
		- env
		- pip3 list
		- python3 -m pytest -v -n $NUM_CORES --reruns 2 --cov-report html --cov-report xml --cov=. --junitxml=report.xml tests/test_vec.py tests/test_random.py tests/test_half_precision.py
		- python3 -m coverage xml
		tags:
		- docker
		- multiarch
		artifacts:
		when: always
		paths:
		- coverage_report
		reports:
		coverage_report:
		coverage_format: cobertura
		path: coverage.xml
		junit: report.xml

		minimal-conda:
		stage: test
		except:
		arm64v8:
		extends: .multiarch_template
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/arm64
		variables:
		QEMU_CPU: "cortex-a76"
		before_script:
		- *multiarch_before_script

		ppc64le:
		extends: .multiarch_template
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/ppc64le
		before_script:
		- *multiarch_before_script

		arm64v9:
		# SVE support is still unreliable in GCC 13 (incorrect code for fixed-width vectors, internal compiler errors).
		# For half precision Clang is necessary
		extends: .multiarch_template
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/arm64
		variables:
		- $ENABLE_NIGHTLY_BUILDS
		# disable pointer authentication to speed up emulation by 3x
		QEMU_CPU: "max,pauth-impdef=on"
		before_script:
		- *multiarch_before_script
		- sed -i s/g\+\+/clang++/g ~/.config/pystencils/config.json

		riscv64:
		# RISC-V vector extension support is incomplete in GCC 13.
		extends: .multiarch_template
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/riscv64
		variables:
		# explicitly set SIMD as detection requires QEMU >= 8.1
		PYSTENCILS_SIMD: "rvv"
		QEMU_CPU: "rv64,v=true,zicboz=true"
		before_script:
		- *multiarch_before_script
		- sed -i 's/march=native/march=rv64imfdvzicboz/g' ~/.config/pystencils/config.json
		- sed -i s/g\+\+/clang++/g ~/.config/pystencils/config.json

		minimal-conda:
		stage: pretest
		extends: .every-commit
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/minimal_conda
		before_script:
		- pip install -e .
		script:
		- python setup.py quicktest
		- python quicktest.py
		tags:
		- docker
		- cuda


		minimal-sympy-master:
		stage: test
		except:
		variables:
		- $ENABLE_NIGHTLY_BUILDS
		extends: .every-commit
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/minimal_conda
		before_script:
		- pip install -e .
		script:
		- python -m pip install --upgrade git+https://github.com/sympy/sympy.git
		- python setup.py quicktest
		- python quicktest.py
		allow_failure: true
		tags:
		- docker
		- cuda


		pycodegen-integration:
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full
		stage: test
		when: manual
		allow_failure: true
		script:
		- git clone https://gitlab-ci-token:${CI_JOB_TOKEN}@i10git.cs.fau.de/pycodegen/pycodegen.git
		- cd pycodegen
		@@ -118,50 +257,89 @@ pycodegen-integration:
		- git fetch test
		- git reset --hard $CI_COMMIT_SHA
		- cd ..
		- export PYTHONPATH=`pwd`/pystencils:`pwd`/lbmpy:`pwd`/pygrandchem:`pwd`/pystencils_walberla:`pwd`/lbmpy_walberla
		- pip install -e pystencils/
		- pip install -e lbmpy/
		- cmake --version
		- ./install_walberla.sh
		- export NUM_CORES=$(nproc --all)
		- mkdir -p ~/.config/matplotlib
		- echo "backend:template" > ~/.config/matplotlib/matplotlibrc
		- cd pystencils
		- py.test -v -n $NUM_CORES .
		- py.test -v -n $NUM_CORES --junitxml=report.xml .
		- cd ../lbmpy
		- py.test -v -n $NUM_CORES .
		- cd ../pygrandchem
		- py.test -v -n $NUM_CORES .
		- py.test -v -n $NUM_CORES --junitxml=report.xml .
		- cd ../walberla/build/
		- make CodegenJacobiCPU CodegenJacobiGPU MicroBenchmarkGpuLbm LbCodeGenerationExample
		- make -j $NUM_CORES CodegenJacobiCPU CodegenJacobiGPU CodegenPoissonCPU CodegenPoissonGPU MicroBenchmarkGpuLbm LbCodeGenerationExample
		- make -j $NUM_CORES multiphaseCPU multiphaseGPU FluctuatingMRT FlowAroundSphereCodeGen FieldLayoutAndVectorizationTest GeneratedOutflowBC
		- cd apps/benchmarks/UniformGridGPU
		- make -j $NUM_CORES
		- cd ../UniformGridCPU
		- make -j $NUM_CORES
		tags:
		- docker
		- cuda
		- cuda11
		- AVX
		artifacts:
		when: always
		reports:
		junit: pycodegen/*/report.xml


		# -------------------- Scheduled Tasks --------------------------------------------------------------------------


		# Nightly test against the latest (pre-release) version of SymPy published on PyPI
		nightly-sympy:
		stage: nightly
		needs: []
		extends: .scheduled
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/latest_python
		before_script:
		- pip install -e .
		- pip install --upgrade --pre sympy
		script:
		- env
		- pip list
		- export NUM_CORES=$(nproc --all)
		- mkdir -p ~/.config/matplotlib
		- echo "backend:template" > ~/.config/matplotlib/matplotlibrc
		- mkdir public
		- pytest -v -n $NUM_CORES -m "not longrun" --junitxml=report.xml
		tags:
		- docker
		- AVX
		- cuda
		artifacts:
		when: always
		reports:
		junit: report.xml

		# -------------------- Linter & Documentation --------------------------------------------------------------------------


		flake8-lint:
		stage: test
		except:
		variables:
		- $ENABLE_NIGHTLY_BUILDS
		stage: pretest
		extends: .every-commit
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full
		script:
		- flake8 pystencils
		- flake8 src/pystencils
		tags:
		- docker
		- cuda


		build-documentation:
		stage: test
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full
		stage: docs
		extends: .every-commit
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/documentation
		needs: []
		before_script:
		- pip install -e .
		script:
		- export PYTHONPATH=`pwd`
		- mkdir html_doc
		- sphinx-build -b html doc html_doc
		- sphinx-build -W -b html doc html_doc
		tags:
		- docker
		- cuda
		artifacts:
		paths:
		- html_doc
		@@ -169,7 +347,9 @@ build-documentation:

		pages:
		image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full
		extends: .every-commit-master
		stage: deploy
		needs: ["tests-and-coverage", "build-documentation"]
		script:
		- ls -l
		- mv coverage_report html_doc
		@@ -179,5 +359,3 @@ pages:
		- public
		tags:
		- docker
		only:
		- master@pycodegen/pystencils

AUTHORS.txt

+6 −5

Original line number	Diff line number	Diff line
		@@ -3,12 +3,13 @@ Contributors:
		-------------

		- Martin Bauer <martin.bauer@fau.de>
		- Markus Holzer <markus.holzer@fau.de>
		- Stephan Seitz <stephan.seitz@fau.de>
		- Michael Kuron <mkuron@icp.uni-stuttgart.de>
		- Jan Hönig <jan.hoenig@fau.de>
		- Nils Kohl <nils.kohl@fau.de>
		- Julian Hammer <julian.hammer@fau.de>
		- Christian Godenschwager <christian.godenschwager@fau.de>
		- Markus Holzer <markus.holzer@fau.de>
		- Michael Kuron <mkuron@icp.uni-stuttgart.de>
		- Nils Kohl <nils.kohl@fau.de>
		- Frederik Hennig <frederik.hennig@fau.de>
		- Dominik Ernst <dominik.ernst@fau.de>
		- João Victor Tozatti Risso <joaovictortr@protonmail.com>
		- Christian Godenschwager <christian.godenschwager@fau.de>
		- Dominik Thoennes <dominik.thoennes@fau.de>

CHANGELOG.md

0 → 100644

+7 −0

Original line number	Diff line number	Diff line
		# Change Log

		## Unreleased

		### Removed
		* LLVM backend because it was not used much and not good integrated in pystencils.
		* OpenCL backend because it was not used much and not good integrated in pystencils.

CONTRIBUTING.md

0 → 100644

+87 −0

Original line number	Diff line number	Diff line
		# Contributing

		Contributions to pystencils are always welcome, and they are greatly appreciated!
		A list of open problems can be found [here]( https://i10git.cs.fau.de/pycodegen/pystencils/-/issues).
		Of course, it is also always appreciated to bring own ideas and problems to the community!


		Please submit all contributions to the official [GitLab repository](https://i10git.cs.fau.de/pycodegen/pystencils) in the form of a Merge Request. Please do not submit git diffs or files containing the changes.
		There also exists a GitHub repository, which is only a mirror to the GitLab repository. Contributions to the GitHub repository are not considered.

		`pystencils` is an open-source python package under the license of AGPL3. Thus we consider the act of contributing to the code by submitting a Merge Request as the "Sign off" or agreement to the AGPL3 license.

		You can contribute in many different ways:

		## Types of Contributions

		### Report Bugs

		Report bugs at [https://i10git.cs.fau.de/pycodegen/pystencils/-/issues](https://i10git.cs.fau.de/pycodegen/pystencils/-/issues).

		For pystencils, it is often necessary to provide the python and [SymPy](https://www.sympy.org/en/index.html) versions used and hardware information like the
		processor architecture and the compiler version used to compile the generated kernels.

		### Fix Issues

		Look through the GitLab issues. Different tags are indicating the status of the issues.
		The "bug" tag indicates problems with pystencils, while the "feature" tag shows ideas that should be added in the future.

		### Write Documentation

		The documentation of pystencils can be found [here](https://pycodegen.pages.i10git.cs.fau.de/pystencils/). Jupyter notebooks are used to provide an
		interactive start to pystencils. It is always appreciated if new document notebooks are provided
		since this helps others a lot.

		## Get Started!

		Ready to contribute? Here is how to set up `pystencils` for local development.

		1. Fork the `pystencils` repo on GitLab.
		2. Clone your fork locally:
		```bash
		$ git clone https://i10git.cs.fau.de/your-name/pystencils
		```
		3. Install your local copy into a virtualenv. It is also recommended to use anaconda or miniconda to manage the python environments.
		```bash
		$ mkvirtualenv pystencils
		$ cd pystencils/
		$ pip install -e .
		```
		4. Create a branch for local development:
		```bash
		$ git checkout -b name-of-your-bugfix-or-feature
		```
		Now you can make your changes locally.

		5. When you're done making changes, check that your changes pass flake8 and the
		tests
		```bash
		$ flake8 pystencils
		$ py.test -v -n $NUM_CORES -m "not longrun" .

		```

		To get all packages needed for development, a requirements list can be found [here](https://i10git.cs.fau.de/pycodegen/pycodegen/-/blob/master/conda_environment_dev.yml). This includes flake8 and pytest.

		6. Commit your changes and push your branch to GitHub::
		```bash
		$ git add .
		$ git commit -m "Your detailed description of your changes."
		$ git push origin name-of-your-bugfix-or-feature
		```
		7. Submit a Merge Request on GitLab.

		## Merge Request Guidelines

		Before you submit a Merge Request, check that it meets these guidelines:

		1. All functionality that is implemented through this Merge Request should be covered by unit tests. These are implemented in `pystencil_tests`
		2. If the Merge Request adds functionality, the docs should be updated. Put your new functionality into a function with a docstring.
		3. If you have a maintainer status for `pystencils`, you can merge Merge Requests to the master branch. However, every Merge Request needs to be reviewed by another developer. Thus it is not allowed to merge a Merge Request, which is submitted by oneself.

		## Tips

		To run a subset of tests:
		```bash
		$ py.test my_test.py
		```
		No newline at end of file

MANIFEST.in

+3 −3

Original line number	Diff line number	Diff line
		include README.md
		include COPYING.txt
		include RELEASE-VERSION
		include AUTHORS.txt
		include CONTRIBUTING.md
		include CHANGELOG.md

README.md

+28 −14

Original line number	Diff line number	Diff line
		@@ -2,16 +2,16 @@ pystencils
		==========

		[![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/mabau/pystencils/master?filepath=doc%2Fnotebooks)
		[![Docs](https://img.shields.io/badge/read-the_docs-brightgreen.svg)](http://pycodegen.pages.walberla.net/pystencils)
		[![Docs](https://img.shields.io/badge/read-the_docs-brightgreen.svg)](https://pycodegen.pages.i10git.cs.fau.de/pystencils)
		[![pypi-package](https://badge.fury.io/py/pystencils.svg)](https://badge.fury.io/py/pystencils)
		[![pipeline status](https://i10git.cs.fau.de/pycodegen/pystencils/badges/master/pipeline.svg)](https://i10git.cs.fau.de/pycodegen/pystencils/commits/master)
		[![coverage report](https://i10git.cs.fau.de/pycodegen/pystencils/badges/master/coverage.svg)](http://pycodegen.pages.walberla.net/pystencils/coverage_report)
		[![coverage report](https://i10git.cs.fau.de/pycodegen/pystencils/badges/master/coverage.svg)](http://pycodegen.pages.i10git.cs.fau.de/pystencils/coverage_report)

		Run blazingly fast stencil codes on numpy arrays.

		pystencils uses sympy to define stencil operations, that can be executed on numpy arrays.
		Exploiting the stencil structure makes pystencils run faster than normal numpy code and even as Cython and numba,
		[as demonstrated in this notebook](http://pycodegen.pages.walberla.net/pystencils/notebooks/demo_benchmark.html).
		[as demonstrated in this notebook](https://pycodegen.pages.i10git.cs.fau.de/pystencils/notebooks/demo_benchmark.html).


		Here is a code snippet that computes the average of neighboring cells:
		@@ -33,15 +33,15 @@ kernel(f=f_arr, g=g_arr)
		It comes with automatic finite difference discretization for PDEs:

		```python
		import pystencils as ps
		import sympy as sp

		c, v = ps.fields("c, v(2): [2D]")
		adv_diff_pde = ps.fd.transient(c) - ps.fd.diffusion(c, sp.symbols("D")) + ps.fd.advection(c, v)
		discretize = ps.fd.Discretization2ndOrder(dx=1, dt=0.01)
		discretization = discretize(adv_diff_pde)
		```

		Look at the [documentation](http://pycodegen.pages.walberla.net/pystencils) to learn more.


		Installation
		------------

		@@ -52,12 +52,10 @@ pip install pystencils[interactive]
		Without `[interactive]` you get a minimal version with very little dependencies.

		All options:
		- `gpu`: use this if an Nvidia GPU is available and CUDA is installed
		- `opencl`: basic OpenCL support (experimental)
		- `gpu`: use this if an NVIDIA or AMD GPU is available and CUDA or ROCm is installed
		- `alltrafos`: pulls in additional dependencies for loop simplification e.g. libisl
		- `bench_db`: functionality to store benchmark result in object databases
		- `interactive`: installs dependencies to work in Jupyter including image I/O, plotting etc.
		- `autodiff`: enable derivation of adjoint kernels and generation of Torch/Tensorflow operations
		- `doc`: packages to build documentation

		Options can be combined e.g.
		@@ -65,9 +63,25 @@ Options can be combined e.g.
		pip install pystencils[interactive, gpu, doc]
		```

		pystencils is also fully compatible with Windows machines. If working with visual studio and cupy makes sure to run example files first to ensure that cupy can find the compiler's executable.

		Documentation
		-------------

		Read the docs [here](http://pycodegen.pages.walberla.net/pystencils) and
		check out the Jupyter notebooks in `doc/notebooks`.
		Read the docs [here](https://pycodegen.pages.i10git.cs.fau.de/pystencils) and
		check out the Jupyter notebooks in `doc/notebooks`. The Changelog of pystencils can be found [here](https://i10git.cs.fau.de/pycodegen/pystencils/-/blob/master/CHANGELOG.md).

		Authors
		-------

		Many thanks go to the [contributors](https://i10git.cs.fau.de/pycodegen/pystencils/-/blob/master/AUTHORS.txt) of pystencils.

		### Please cite us

		If you use pystencils in a publication, please cite the following articles:

		Overview:
		- M. Bauer et al, Code Generation for Massively Parallel Phase-Field Simulations. Association for Computing Machinery, 2019. https://doi.org/10.1145/3295500.3356186

		Performance Modelling:
		- D. Ernst et al, Analytical performance estimation during code generation on modern GPUs. Journal of Parallel and Distributed Computing, 2023. https://doi.org/10.1016/j.jpdc.2022.11.003

binder/environment.yml

+3 −3

Original line number	Diff line number	Diff line
		@@ -7,14 +7,14 @@
		# conda env create -f conda_environment_user.yml
		# . activate pystencils
		#
		# If you have CUDA installed and want to use your GPU, uncomment the last line to install pycuda
		# If you have CUDA or ROCm installed and want to use your GPU, uncomment the last line to install cupy
		#
		# ----------------------------------------------------------------------------------------------------------------------

		name: pystencils
		dependencies:
		# Basic dependencies:
		- python >= 3.6
		- python >= 3.8
		- numpy
		- sympy >= 1.1
		- appdirs # to find default cache directory on each platform
		@@ -32,4 +32,4 @@ dependencies:
		- ipy_table # HTML tables for jupyter notebooks
		- pyevtk # VTK output for serial simulations
		- blitzdb # file-based No-SQL database to store simulation results
		#- pycuda # add this if you have CUDA installed
		#- cupy # add this if you have CUDA or ROCm installed

conftest.py

+47 −32

Original line number	Diff line number	Diff line
		import os
		import pytest
		import tempfile
		import runpy
		import sys
		import tempfile
		import warnings

		import nbformat
		import pytest
		from nbconvert import PythonExporter

		from pystencils.boundaries.createindexlist import * # NOQA
		# Trigger config file reading / creation once - to avoid race conditions when multiple instances are creating it
		# at the same time
		from pystencils.cpu import cpujit
		@@ -15,12 +20,17 @@ try:
		pyximport.install(language_level=3)
		except ImportError:
		pass
		from pystencils.boundaries.createindexlistcython import * # NOQA


		SCRIPT_FOLDER = os.path.dirname(os.path.realpath(__file__))
		sys.path.insert(0, os.path.abspath('pystencils'))

		# the Ubuntu pipeline uses an older version of pytest which uses deprecated functionality.
		# This leads to many warinings in the test and coverage pipeline.
		pytest_numeric_version = [int(x, 10) for x in pytest.__version__.split('.')]
		pytest_numeric_version.reverse()
		pytest_version = sum(x * (100 ** i) for i, x in enumerate(pytest_numeric_version))


		def add_path_to_ignore(path):
		if not os.path.exists(path):
		@@ -29,46 +39,49 @@ def add_path_to_ignore(path):
		collect_ignore += [os.path.join(SCRIPT_FOLDER, path, f) for f in os.listdir(os.path.join(SCRIPT_FOLDER, path))]


		collect_ignore = [os.path.join(SCRIPT_FOLDER, "doc", "conf.py")]
		add_path_to_ignore('pystencils_tests/benchmark')
		collect_ignore = [os.path.join(SCRIPT_FOLDER, "doc", "conf.py"),
		os.path.join(SCRIPT_FOLDER, "src", "pystencils", "opencl", "opencl.autoinit")]
		add_path_to_ignore('tests/benchmark')
		add_path_to_ignore('_local_tmp')


		collect_ignore += [os.path.join(SCRIPT_FOLDER, "pystencils/autodiff.py")]

		try:
		import pycuda
		import cupy
		except ImportError:
		collect_ignore += [os.path.join(SCRIPT_FOLDER, "pystencils/pystencils_tests/test_cudagpu.py")]
		add_path_to_ignore('pystencils/gpucuda')
		collect_ignore += [os.path.join(SCRIPT_FOLDER, "tests/test_gpu.py")]
		add_path_to_ignore('src/pystencils/gpu')

		try:
		import llvmlite
		import waLBerla
		except ImportError:
		collect_ignore += [os.path.join(SCRIPT_FOLDER, 'pystencils_tests/backends/llvm.py')]
		add_path_to_ignore('pystencils/llvm')
		collect_ignore += [os.path.join(SCRIPT_FOLDER, "tests/test_aligned_array.py"),
		os.path.join(SCRIPT_FOLDER, "tests/test_datahandling_parallel.py"),
		os.path.join(SCRIPT_FOLDER, "doc/notebooks/03_tutorial_datahandling.ipynb"),
		os.path.join(SCRIPT_FOLDER, "src/pystencils/datahandling/parallel_datahandling.py"),
		os.path.join(SCRIPT_FOLDER, "tests/test_small_block_benchmark.ipynb")]

		try:
		import kerncraft
		import blitzdb
		except ImportError:
		collect_ignore += [os.path.join(SCRIPT_FOLDER, "pystencils_tests/test_kerncraft_coupling.py"),
		os.path.join(SCRIPT_FOLDER, "pystencils_tests/benchmark/benchmark.py")]
		add_path_to_ignore('pystencils/kerncraft_coupling')
		add_path_to_ignore('src/pystencils/runhelper')
		collect_ignore += [os.path.join(SCRIPT_FOLDER, "tests/test_parameterstudy.py")]
		collect_ignore += [os.path.join(SCRIPT_FOLDER, "tests/test_json_serializer.py")]

		try:
		import waLBerla
		import islpy
		except ImportError:
		collect_ignore += [os.path.join(SCRIPT_FOLDER, "pystencils_tests/test_aligned_array.py"),
		os.path.join(SCRIPT_FOLDER, "pystencils_tests/test_datahandling_parallel.py"),
		os.path.join(SCRIPT_FOLDER, "doc/notebooks/03_tutorial_datahandling.ipynb"),
		os.path.join(SCRIPT_FOLDER, "pystencils/datahandling/parallel_datahandling.py"),
		os.path.join(SCRIPT_FOLDER, "pystencils_tests/test_small_block_benchmark.ipynb")]
		collect_ignore += [os.path.join(SCRIPT_FOLDER, "src/pystencils/integer_set_analysis.py")]

		try:
		import blitzdb
		import graphviz
		except ImportError:
		add_path_to_ignore('pystencils/runhelper')
		collect_ignore += [os.path.join(SCRIPT_FOLDER, "src/pystencils/backends/dot.py")]
		collect_ignore += [os.path.join(SCRIPT_FOLDER, "doc/notebooks/01_tutorial_getting_started.ipynb")]

		try:
		import pyevtk
		except ImportError:
		collect_ignore += [os.path.join(SCRIPT_FOLDER, "src/pystencils/datahandling/vtk.py")]

		collect_ignore += [os.path.join(SCRIPT_FOLDER, 'setup.py')]

		@@ -78,10 +91,6 @@ for root, sub_dirs, files in os.walk('.'):
		collect_ignore.append(f)


		import nbformat
		from nbconvert import PythonExporter


		class IPythonMockup:
		def run_line_magic(self, args, *kwargs):
		pass
		@@ -128,12 +137,15 @@ class IPyNbFile(pytest.File):
		exporter.exclude_markdown = True
		exporter.exclude_input_prompt = True

		notebook_contents = self.fspath.open()
		notebook_contents = self.fspath.open(encoding='utf-8')

		with warnings.catch_warnings():
		warnings.filterwarnings("ignore", "IPython.core.inputsplitter is deprecated")
		notebook = nbformat.read(notebook_contents, 4)
		code, _ = exporter.from_notebook_node(notebook)
		if pytest_version >= 50403:
		yield IPyNbTest.from_parent(name=self.name, parent=self, code=code)
		else:
		yield IPyNbTest(self.name, self, code)

		def teardown(self):
		@@ -143,4 +155,7 @@ class IPyNbFile(pytest.File):
		def pytest_collect_file(path, parent):
		glob_exprs = ["demo.ipynb", "tutorial.ipynb", "test_*.ipynb"]
		if any(path.fnmatch(g) for g in glob_exprs):
		if pytest_version >= 50403:
		return IPyNbFile.from_parent(fspath=path, parent=parent)
		else:
		return IPyNbFile(path, parent)

doc/conf.py

100644 → 100755

+13 −8

Original line number	Diff line number	Diff line
		@@ -4,10 +4,11 @@
		import datetime
		import sphinx_rtd_theme
		import os
		import re
		import sys

		sys.path.insert(0, os.path.abspath('.'))
		from version_from_git import version_number_from_git
		import pystencils

		extensions = [
		'sphinx.ext.autodoc',
		@@ -25,11 +26,14 @@ templates_path = ['_templates']
		source_suffix = '.rst'
		master_doc = 'index'

		copyright = '{}, Martin Bauer'.format(datetime.datetime.now().year)
		author = 'Martin Bauer'
		version = version_number_from_git()
		release = version_number_from_git()
		language = None
		copyright = f'{datetime.datetime.now().year}, Martin Bauer, Markus Holzer, Frederik Hennig'
		author = 'Martin Bauer, Markus Holzer, Frederik Hennig'
		# The short X.Y version (including .devXXXX, rcX, b1 suffixes if present)
		version = re.sub(r'(\d+\.\d+)\.\d+(.*)', r'\1\2', pystencils.__version__)
		version = re.sub(r'(\.dev\d+).*?$', r'\1', version)
		# The full version, including alpha/beta/rc tags.
		release = pystencils.__version__
		language = 'en'
		exclude_patterns = ['_build', 'Thumbs.db', '.DS_Store', '**.ipynb_checkpoints']
		default_role = 'any'
		pygments_style = 'sphinx'
		@@ -47,13 +51,14 @@ nbsphinx_execute = 'never'
		nbsphinx_codecell_lexer = 'python3'

		# Example configuration for intersphinx: refer to the Python standard library.
		intersphinx_mapping = {'python': ('https://docs.python.org/3.6', None),
		intersphinx_mapping = {'python': ('https://docs.python.org/3.8', None),
		'numpy': ('https://docs.scipy.org/doc/numpy/', None),
		'matplotlib': ('https://matplotlib.org/', None),
		'sympy': ('https://docs.sympy.org/latest/', None),
		}

		autodoc_member_order = 'bysource'
		bibtex_bibfiles = ['sphinx/pystencils.bib']

		project = 'pystencils'
		html_logo = "img/logo.png"
		html_logo = 'img/logo.png'

doc/index.rst

+1 −0

Original line number	Diff line number	Diff line
		@@ -14,5 +14,6 @@ pystencils can help you to generate blazingly fast code for image processing, nu

		.. image:: /img/pystencils_arch_block_diagram.svg
		:height: 450px
		:align: center

doc/notebooks/01_tutorial_getting_started.ipynb

+431 −448

File changed.

File size exceeds preview limit.

View original file

View changed file