Skip to content
Snippets Groups Projects

Compare revisions

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

Source

Select target project
No results found
Select Git revision
  • 66-absolute-access-is-probably-not-copied-correctly-after-_eval_subs
  • const_fix
  • fhennig/v2.0-deprecations
  • fma
  • gpu_bufferfield_fix
  • gpu_liveness_opts
  • holzer-master-patch-46757
  • hyteg
  • improved_comm
  • master
  • nhr-course-week
  • target_dh_refactoring
  • v2.0-dev
  • vectorization_sqrt_fix
  • zikeliml/124-rework-tutorials
  • zikeliml/Task-96-dotExporterForAST
  • last/Kerncraft
  • last/LLVM
  • last/OpenCL
  • release/0.2.1
  • release/0.2.10
  • release/0.2.11
  • release/0.2.12
  • release/0.2.13
  • release/0.2.14
  • release/0.2.15
  • release/0.2.2
  • release/0.2.3
  • release/0.2.4
  • release/0.2.6
  • release/0.2.7
  • release/0.2.8
  • release/0.2.9
  • release/0.3.0
  • release/0.3.1
  • release/0.3.2
  • release/0.3.3
  • release/0.3.4
  • release/0.4.0
  • release/0.4.1
  • release/0.4.2
  • release/0.4.3
  • release/0.4.4
  • release/1.0
  • release/1.0.1
  • release/1.1
  • release/1.1.1
  • release/1.2
  • release/1.3
  • release/1.3.1
  • release/1.3.2
  • release/1.3.3
  • release/1.3.4
  • release/1.3.5
  • release/1.3.6
  • release/1.3.7
  • release/2.0.dev0
57 results

Target

Select target project
  • anirudh.jonnalagadda/pystencils
  • hyteg/pystencils
  • jbadwaik/pystencils
  • jngrad/pystencils
  • itischler/pystencils
  • ob28imeq/pystencils
  • hoenig/pystencils
  • Bindgen/pystencils
  • hammer/pystencils
  • da15siwa/pystencils
  • holzer/pystencils
  • alexander.reinauer/pystencils
  • ec93ujoh/pystencils
  • Harke/pystencils
  • seitz/pystencils
  • pycodegen/pystencils
16 results
Select Git revision
  • VoF
  • abs
  • compare_fix
  • const_fix
  • gpu_liveness_opts
  • hyteg
  • improved_comm
  • jan_test
  • master
  • noFlux
  • philox-simd
  • target_dh_refactoring
  • test_martin
  • test_martin2
  • last/Kerncraft
  • last/LLVM
  • last/OpenCL
  • release/0.2.1
  • release/0.2.10
  • release/0.2.11
  • release/0.2.12
  • release/0.2.13
  • release/0.2.14
  • release/0.2.15
  • release/0.2.2
  • release/0.2.3
  • release/0.2.4
  • release/0.2.6
  • release/0.2.7
  • release/0.2.8
  • release/0.2.9
  • release/0.3.0
  • release/0.3.1
  • release/0.3.2
  • release/0.3.3
  • release/0.3.4
  • release/0.4.0
  • release/0.4.1
  • release/0.4.2
  • release/0.4.3
  • release/0.4.4
  • release/1.0
  • release/1.0.1
  • release/1.1
  • release/1.1.1
  • release/1.2
  • release/1.3
  • release/1.3.1
  • release/1.3.2
  • release/1.3.3
50 results
Show changes
Commits on Source (83)
Hide whitespace changes
Inline Side-by-side
Showing
with 365 additions and 596 deletions
.flake8
View file @ 5600b6b6
[flake8]
max-line-length=120
exclude=pystencils/jupyter.py,
pystencils/plot.py
pystencils/session.py
exclude=src/pystencils/jupyter.py,
src/pystencils/plot.py
src/pystencils/session.py
ignore = W293 W503 W291 C901 E741
.gitattributes
View file @ 5600b6b6
pystencils/_version.py export-subst
src/pystencils/_version.py export-subst
.gitignore
View file @ 5600b6b6
......@@ -5,7 +5,7 @@ __pycache__
*.vti
/build
/dist
/*.egg-info
*.egg-info
.cache
_build
/html_doc
......@@ -15,12 +15,12 @@ _build
_local_tmp
RELEASE-VERSION
test-report
pystencils/boundaries/createindexlistcython.c
pystencils/boundaries/createindexlistcython.*.so
pystencils_tests/tmp
pystencils_tests/var
pystencils_tests/kerncraft_inputs/.2d-5pt.c_kerncraft/
pystencils_tests/kerncraft_inputs/.3d-7pt.c_kerncraft/
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/
......
.gitlab-ci.yml
View file @ 5600b6b6
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: pretest
except:
variables:
- $ENABLE_NIGHTLY_BUILDS
image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full
extends: .every-commit
image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full:cupy12.3
before_script:
- pip install -e .
script:
- pip install sympy --upgrade
- env
- pip list
- export NUM_CORES=$(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 xml --cov-report term --cov=. -m "not longrun" --html test-report/index.html --junitxml=report.xml
- python3 -m coverage xml
- pytest -v -n $NUM_CORES --cov-report html --cov-report xml --cov-report term --cov=. -m "not longrun" --html test-report/index.html --junitxml=report.xml
- python -m coverage xml
tags:
- docker
- cuda11
......@@ -44,9 +62,11 @@ tests-and-coverage-with-longrun:
stage: test
when: manual
allow_failure: true
image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full
script:
image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full:cupy12.3
before_script:
- pip install sympy --upgrade
- pip install -e .
script:
- env
- pip list
- export NUM_CORES=$(nproc --all)
......@@ -62,13 +82,14 @@ tests-and-coverage-with-longrun:
# pipeline with latest python version
latest-python:
stage: test
except:
variables:
- $ENABLE_NIGHTLY_BUILDS
extends: .every-commit
image: i10git.cs.fau.de:5005/pycodegen/pycodegen/latest_python
before_script:
- pip install -e .
script:
- env
- pip list
- pip install -e .
- export NUM_CORES=$(nproc --all)
- mkdir -p ~/.config/matplotlib
- echo "backend:template" > ~/.config/matplotlib/matplotlibrc
......@@ -86,9 +107,6 @@ latest-python:
# Minimal tests in windows environment
#minimal-windows:
# stage: test
# except:
# variables:
# - $ENABLE_NIGHTLY_BUILDS
# tags:
# - win
# script:
......@@ -102,23 +120,19 @@ latest-python:
ubuntu:
stage: test
except:
variables:
- $ENABLE_NIGHTLY_BUILDS
extends: .every-commit
image: i10git.cs.fau.de:5005/pycodegen/pycodegen/ubuntu
before_script:
- apt-get -y remove python3-sympy
- ln -s /usr/include/locale.h /usr/include/xlocale.h
- pip3 install `grep -Eo 'sympy[>=]+[0-9\.]+' setup.py | sed 's/>/=/g'`
# - pip3 install `grep -Eo 'sympy[>=]+[0-9\.]+' setup.py | sed 's/>/=/g'`
- pip3 install -e .
script:
- export NUM_CORES=$(nproc --all)
- mkdir -p ~/.config/matplotlib
- echo "backend:template" > ~/.config/matplotlib/matplotlibrc
- sed -i 's/--doctest-modules //g' pytest.ini
- env
- pip3 list
- pytest-3 -v -n $NUM_CORES -m "not longrun" --junitxml=report.xml
- pip list
- pytest -v -n $NUM_CORES -m "not longrun" --junitxml=report.xml
tags:
- docker
- cuda11
......@@ -130,10 +144,11 @@ ubuntu:
.multiarch_template:
stage: test
except:
variables:
- $ENABLE_NIGHTLY_BUILDS
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:
......@@ -143,13 +158,19 @@ ubuntu:
- sed -i 's/--doctest-modules //g' pytest.ini
- env
- pip3 list
- pytest-3 -v -n $NUM_CORES --junitxml=report.xml pystencils_tests/test_*vec*.py pystencils_tests/test_random.py pystencils_tests/test_half_precision.py
- python3 -m pytest -v -n $NUM_CORES --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
- AVX
artifacts:
when: always
paths:
- coverage_report
reports:
coverage_report:
coverage_format: cobertura
path: coverage.xml
junit: report.xml
arm64v8:
......@@ -175,32 +196,30 @@ arm64v9:
image: i10git.cs.fau.de:5005/pycodegen/pycodegen/arm64
before_script:
- *multiarch_before_script
- sed -i s/march=native/march=armv8-a+sve/g ~/.config/pystencils/config.json
- sed -i s/march=native/march=armv9-a+sve2+sme/g ~/.config/pystencils/config.json
- sed -i s/g\+\+/clang++/g ~/.config/pystencils/config.json
riscv64:
# RISC-V vector extension are currently not supported by GCC.
# Also, the image is built without the libomp package which is not yet available on Ubuntu.
extends: .multiarch_template
image: i10git.cs.fau.de:5005/pycodegen/pycodegen/riscv64
variables:
# explicitly set SIMD as detection does not appear to work on QEMU
# explicitly set SIMD as detection requires QEMU >= 8.1
PYSTENCILS_SIMD: "rvv"
QEMU_CPU: "rv64,v=true"
QEMU_CPU: "rv64,v=true,zicboz=true"
before_script:
- *multiarch_before_script
- sed -i 's/march=native/march=rv64imfdv/g' ~/.config/pystencils/config.json
- sed -i s/g\+\+/clang++/g ~/.config/pystencils/config.json
- sed -i 's/fopenmp/fopenmp=libgomp -I\/usr\/include\/riscv64-linux-gnu/g' ~/.config/pystencils/config.json
- sed -i 's/march=native/march=rv64imfdvzicboz/g' ~/.config/pystencils/config.json
- sed -i s/g\+\+/clang++-15/g ~/.config/pystencils/config.json
minimal-conda:
stage: pretest
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 setup.py quicktest
- python quicktest.py
tags:
- docker
- cuda
......@@ -208,13 +227,13 @@ minimal-conda:
minimal-sympy-master:
stage: test
except:
variables:
- $ENABLE_NIGHTLY_BUILDS
extends: .every-commit
image: i10git.cs.fau.de:5005/pycodegen/pycodegen/minimal_conda
before_script:
- pip install -e .
script:
- python -m pip install --upgrade git+https://github.com/sympy/sympy.git
- python setup.py quicktest
- python quicktest.py
allow_failure: true
tags:
- docker
......@@ -265,26 +284,57 @@ pycodegen-integration:
reports:
junit: pycodegen/*/report.xml
# -------------------- Scheduled Tasks --------------------------------------------------------------------------
# Nightly test against the latest (pre-release) version of SymPy published on PyPI
nightly-sympy:
stage: nightly
needs: []
extends: .scheduled
image: i10git.cs.fau.de:5005/pycodegen/pycodegen/latest_python
before_script:
- pip install -e .
- pip install --upgrade --pre sympy
script:
- env
- pip list
- export NUM_CORES=$(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: pretest
except:
variables:
- $ENABLE_NIGHTLY_BUILDS
extends: .every-commit
image: i10git.cs.fau.de:5005/pycodegen/pycodegen/full
script:
- flake8 pystencils
- flake8 src/pystencils
tags:
- docker
build-documentation:
stage: test
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
......@@ -297,7 +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
......@@ -307,5 +359,3 @@ pages:
- public
tags:
- docker
only:
- master@pycodegen/pystencils
MANIFEST.in
View file @ 5600b6b6
include README.md
include COPYING.txt
include AUTHORS.txt
include CONTRIBUTING.md
CHANGELOG.md
global-include *.pyx
include versioneer.py
include pystencils/_version.py
include CHANGELOG.md
README.md
View file @ 5600b6b6
......@@ -81,4 +81,7 @@ Many thanks go to the [contributors](https://i10git.cs.fau.de/pycodegen/pystenci
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
\ No newline at end of file
- 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
conftest.py
View file @ 5600b6b6
......@@ -8,7 +8,7 @@ import nbformat
import pytest
from nbconvert import PythonExporter
from pystencils.boundaries.createindexlistcython import * # NOQA
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
......@@ -40,48 +40,48 @@ def add_path_to_ignore(path):
collect_ignore = [os.path.join(SCRIPT_FOLDER, "doc", "conf.py"),
os.path.join(SCRIPT_FOLDER, "pystencils", "opencl", "opencl.autoinit")]
add_path_to_ignore('pystencils_tests/benchmark')
os.path.join(SCRIPT_FOLDER, "src", "pystencils", "opencl", "opencl.autoinit")]
add_path_to_ignore('tests/benchmark')
add_path_to_ignore('_local_tmp')
try:
import cupy
except ImportError:
collect_ignore += [os.path.join(SCRIPT_FOLDER, "pystencils_tests/test_gpu.py")]
add_path_to_ignore('pystencils/gpu')
collect_ignore += [os.path.join(SCRIPT_FOLDER, "tests/test_gpu.py")]
add_path_to_ignore('src/pystencils/gpu')
try:
import waLBerla
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"),
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, "pystencils/datahandling/parallel_datahandling.py"),
os.path.join(SCRIPT_FOLDER, "pystencils_tests/test_small_block_benchmark.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 blitzdb
except ImportError:
add_path_to_ignore('pystencils/runhelper')
collect_ignore += [os.path.join(SCRIPT_FOLDER, "pystencils_tests/test_parameterstudy.py")]
collect_ignore += [os.path.join(SCRIPT_FOLDER, "pystencils_tests/test_json_serializer.py")]
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 islpy
except ImportError:
collect_ignore += [os.path.join(SCRIPT_FOLDER, "pystencils/integer_set_analysis.py")]
collect_ignore += [os.path.join(SCRIPT_FOLDER, "src/pystencils/integer_set_analysis.py")]
try:
import graphviz
except ImportError:
collect_ignore += [os.path.join(SCRIPT_FOLDER, "pystencils/backends/dot.py")]
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, "pystencils/datahandling/vtk.py")]
collect_ignore += [os.path.join(SCRIPT_FOLDER, "src/pystencils/datahandling/vtk.py")]
collect_ignore += [os.path.join(SCRIPT_FOLDER, 'setup.py')]
......
doc/notebooks/02_tutorial_basic_kernels.ipynb
View file @ 5600b6b6
%% Cell type:code id: tags:
``` python
from pystencils.session import *
```
%% Cell type:markdown id: tags:
# Tutorial 02: Basic Kernel generation with *pystencils*
Now that you have an [overview of pystencils](01_tutorial_getting_started.ipynb),
this tutorial shows in more detail how to formulate, optimize and run stencil kernels.
## 1) Kernel Definition
### a) Defining kernels with assignment lists and the `kernel` decorator
*pystencils* gets a symbolic formulation of the kernel. This can be either an `Assignment` or a sequence of `Assignment`s that follow a set of restrictions.
Lets first create a kernel that consists of multiple assignments:
%% Cell type:code id: tags:
``` python
src_arr = np.zeros([20, 30])
dst_arr = np.zeros_like(src_arr)
dst, src = ps.fields(dst=dst_arr, src=src_arr)
```
%% Cell type:code id: tags:
``` python
grad_x, grad_y = sp.symbols("grad_x, grad_y")
symbolic_description = [
ps.Assignment(grad_x, (src[1, 0] - src[-1, 0]) / 2),
ps.Assignment(grad_y, (src[0, 1] - src[0, -1]) / 2),
ps.Assignment(dst[0, 0], grad_x + grad_y),
]
kernel = ps.create_kernel(symbolic_description)
symbolic_description
```
%% Output
$\displaystyle \left[ grad_{x} \leftarrow \frac{{src}_{(1,0)}}{2} - \frac{{src}_{(-1,0)}}{2}, \ grad_{y} \leftarrow \frac{{src}_{(0,1)}}{2} - \frac{{src}_{(0,-1)}}{2}, \ {dst}_{(0,0)} \leftarrow grad_{x} + grad_{y}\right]$
$\displaystyle \left[ grad_{x} \leftarrow_{} \frac{{src}_{(1,0)}}{2} - \frac{{src}_{(-1,0)}}{2}, \ grad_{y} \leftarrow_{} \frac{{src}_{(0,1)}}{2} - \frac{{src}_{(0,-1)}}{2}, \ {dst}_{(0,0)} \leftarrow_{} grad_{x} + grad_{y}\right]$
⎡ src_E src_W src_N src_S ⎤
⎢gradₓ := ───── - ─────, grad_y := ───── - ─────, dst_C := gradₓ + grad_y⎥
⎣ 2 2 2 2 ⎦
%% Cell type:markdown id: tags:
We created subexpressions, using standard sympy symbols on the left hand side, to split the kernel into multiple assignments. Defining a kernel using a list of `Assignment`s is quite tedious and hard to read.
To simplify the formulation of a kernel, *pystencils* offers the `kernel` decorator, that transforms a normal Python function with `@=` assignments into an assignment list that can be passed to `create_kernel`.
%% Cell type:code id: tags:
``` python
@ps.kernel
def symbolic_description_using_function():
grad_x @= (src[1, 0] - src[-1, 0]) / 2
grad_y @= (src[0, 1] - src[0, -1]) / 2
dst[0, 0] @= grad_x + grad_y
symbolic_description_using_function
```
%% Output
$\displaystyle \left[ grad_{x} \leftarrow \frac{{src}_{(1,0)}}{2} - \frac{{src}_{(-1,0)}}{2}, \ grad_{y} \leftarrow \frac{{src}_{(0,1)}}{2} - \frac{{src}_{(0,-1)}}{2}, \ {dst}_{(0,0)} \leftarrow grad_{x} + grad_{y}\right]$
$\displaystyle \left[ grad_{x} \leftarrow_{} \frac{{src}_{(1,0)}}{2} - \frac{{src}_{(-1,0)}}{2}, \ grad_{y} \leftarrow_{} \frac{{src}_{(0,1)}}{2} - \frac{{src}_{(0,-1)}}{2}, \ {dst}_{(0,0)} \leftarrow_{} grad_{x} + grad_{y}\right]$
⎡ src_E src_W src_N src_S ⎤
⎢gradₓ := ───── - ─────, grad_y := ───── - ─────, dst_C := gradₓ + grad_y⎥
⎣ 2 2 2 2 ⎦
%% Cell type:markdown id: tags:
The decorated function can contain any Python code, only the `@=` operator, and the ternary inline `if-else` operator have different meaning.
### b) Ternary 'if' with `Piecewise`
The ternary operator maps to `sympy.Piecewise` functions, that can be used to introduce branching into the kernel. Piecewise defined functions must give a value for every input, i.e. there must be a 'otherwise' clause in the end that is indicated by the condition `True`. Piecewise objects are standard sympy terms that can be integrated into bigger expressions:
%% Cell type:code id: tags:
``` python
sp.Piecewise((1.0, src[0,1] > 0), (0.0, True)) + src[1, 0]
```
%% Output
$\displaystyle {src}_{(1,0)} + \begin{cases} 1.0 & \text{for}\: {src}_{(0,1)} > 0 \\0.0 & \text{otherwise} \end{cases}$
⎛⎧1.0 for src_N > 0⎞
src_E + ⎜⎨ ⎟
⎝⎩0.0 otherwise ⎠
%% Cell type:markdown id: tags:
Piecewise objects are created by the `kernel` decorator for ternary if-else statements.
%% Cell type:code id: tags:
``` python
@ps.kernel
def kernel_with_piecewise():
grad_x @= (src[1, 0] - src[-1, 0]) / 2 if src[-1, 0] > 0 else 0.0
kernel_with_piecewise
```
%% Output
$\displaystyle \left[ grad_{x} \leftarrow \begin{cases} \frac{{src}_{(1,0)}}{2} - \frac{{src}_{(-1,0)}}{2} & \text{for}\: {src}_{(-1,0)} > 0 \\0.0 & \text{otherwise} \end{cases}\right]$
$\displaystyle \left[ grad_{x} \leftarrow_{} \begin{cases} \frac{{src}_{(1,0)}}{2} - \frac{{src}_{(-1,0)}}{2} & \text{for}\: {src}_{(-1,0)} > 0 \\0.0 & \text{otherwise} \end{cases}\right]$
⎡ ⎧src_E src_W ⎤
⎢ ⎪───── - ───── for src_W > 0⎥
⎢gradₓ := ⎨ 2 2 ⎥
⎢ ⎪ ⎥
⎣ ⎩ 0.0 otherwise ⎦
%% Cell type:markdown id: tags:
### c) Assignment level optimizations using `AssignmentCollection`
When the kernels get larger and more complex, it is helpful to organize the list of assignment into a more structured way. The `AssignmentCollection` offers optimizating transformation on a list of assignments. It holds two assignment lists, one for subexpressions and one for the main assignments. Main assignments are typically those that write to an array.
%% Cell type:code id: tags:
``` python
@ps.kernel
def somewhat_longer_dummy_kernel(s):
s.a @= src[0, 1] + src[-1, 0]
s.b @= 2 * src[1, 0] + src[0, -1]
s.c @= src[0, 1] + 2 * src[1, 0] + src[-1, 0] + src[0, -1] - src[0,0]
dst[0, 0] @= s.a + s.b + s.c
ac = ps.AssignmentCollection(main_assignments=somewhat_longer_dummy_kernel[-1:],
subexpressions=somewhat_longer_dummy_kernel[:-1])
ac
```
%% Output
AssignmentCollection: dst_C, <- f(src_N, src_E, src_W, src_C, src_S)
AssignmentCollection: dst_C, <- f(src_C, src_W, src_S, src_N, src_E)
%% Cell type:code id: tags:
``` python
ac.operation_count
```
%% Output
{'adds': 8,
'muls': 2,
'divs': 0,
'sqrts': 0,
'fast_sqrts': 0,
'fast_inv_sqrts': 0,
'fast_div': 0}
%% Cell type:markdown id: tags:
The `pystencils.simp` submodule offers several functions to optimize a collection of assignments.
It also offers functionality to group optimization into strategies and evaluate them.
In this example we reduce the number of operations by reusing existing subexpressions to get rid of two unnecessary floating point additions. For more information about assignment collections and simplifications see the [demo notebook](demo_assignment_collection.ipynb).
%% Cell type:code id: tags:
``` python
opt_ac = ps.simp.subexpression_substitution_in_existing_subexpressions(ac)
opt_ac
```
%% Output
AssignmentCollection: dst_C, <- f(src_N, src_E, src_W, src_C, src_S)
AssignmentCollection: dst_C, <- f(src_C, src_W, src_S, src_N, src_E)
%% Cell type:code id: tags:
``` python
opt_ac.operation_count
```
%% Output
{'adds': 6,
'muls': 1,
'divs': 0,
'sqrts': 0,
'fast_sqrts': 0,
'fast_inv_sqrts': 0,
'fast_div': 0}
%% Cell type:markdown id: tags:
### d) Ghost layers and iteration region
When creating a kernel with neighbor accesses, *pystencils* automatically restricts the iteration region, such that all accesses are safe.
%% Cell type:code id: tags:
``` python
kernel = ps.create_kernel(ps.Assignment(dst[0,0], src[2, 0] + src[-1, 0]))
ps.show_code(kernel)
```
%% Output
%% Cell type:markdown id: tags:
When no additional ghost layer information is given, *pystencils* looks at all neighboring field accesses and introduces the required number of ghost layers **for all directions**. In the example above the largest neighbor accesses was ``src[2, 0]``, so theoretically we would need 2 ghost layers only the the end of the x coordinate.
By default *pystencils* introduces 2 ghost layers at all borders of the domain. The next cell shows how to change this behavior. Be careful with manual ghost layer specification, wrong values may lead to SEGFAULTs.
%% Cell type:code id: tags:
``` python
gl_spec = [(0, 2), # 0 ghost layers at the left, 2 at the right border
(1, 0)] # 1 ghost layer at the lower y, one at the upper y coordinate
kernel = ps.create_kernel(ps.Assignment(dst[0,0], src[2, 0] + src[-1, 0]), ghost_layers=gl_spec)
ps.show_code(kernel)
```
%% Output
%% Cell type:markdown id: tags:
## 2 ) Restrictions
### a) Independence Restriction
*pystencils* only works for kernels where each array element can be updated independently from all other elements. This restriction ensures that the kernels can be easily parallelized and also be run on the GPU. Trying to define kernels where the results depends on the iteration order, leads to a ValueError.
%% Cell type:code id: tags:
``` python
invalid_description = [
ps.Assignment(dst[1, 0], src[1, 0] + src[-1, 0]),
ps.Assignment(dst[0, 0], src[1, 0] - src[-1, 0]),
]
try:
invalid_kernel = ps.create_kernel(invalid_description)
assert False, "Should never be executed"
except ValueError as e:
print(e)
```
%% Output
Field dst is written at two different locations
%% Cell type:markdown id: tags:
The independence restriction makes sure that the kernel can be safely parallelized by checking the following conditions: If a field is modified inside the kernel, it may only be modified at a single spatial position. In that case the field may also only be read at this position. Fields that are not modified may be read at multiple neighboring positions.
Specifically, this rule allows for in-place updates that don't access neighbors.
%% Cell type:code id: tags:
``` python
valid_kernel = ps.create_kernel(ps.Assignment(src[0,0], 2*src[0,0] + 42))
```
%% Cell type:markdown id: tags:
If a field stores multiple values per cell, as in the next example, this restriction only applies for accesses with the same index.
%% Cell type:code id: tags:
``` python
v = ps.fields("v(2): double[2D]")
valid_kernel = ps.create_kernel([ps.Assignment(v[0,0](1), 2*v[0,0](1) + 42),
ps.Assignment(v[0,1](0), 2*v[1,0](0) + 42)])
ps.Assignment(v[0,1](0), 2*v[0,1](0) + 42)])
```
%% Cell type:markdown id: tags:
### b) Static Single Assignment Form
All assignments that don't write to a field must be in SSA form
1. Each sympy symbol may only occur once as a left-hand-side (fields can be written multiple times)
2. A symbol has to be defined before it is used. If it is never defined it is introduced as function parameter
The next cell demonstrates the first SSA restriction:
%% Cell type:code id: tags:
``` python
@ps.kernel
def not_allowed():
a, b = sp.symbols("a b")
a @= src[0, 0]
b @= a + 3
a @= src[-1, 0]
dst[0, 0] @= a + b
try:
ps.create_kernel(not_allowed)
assert False
except ValueError as e:
print(e)
```
%% Output
Assignments not in SSA form, multiple assignments to a
%% Cell type:markdown id: tags:
Also it is not allowed to write a field at the same location
%% Cell type:code id: tags:
``` python
@ps.kernel
def not_allowed():
dst[0, 0] @= src[0, 1] + src[1, 0]
dst[0, 0] @= 2 * dst[0, 0]
try:
ps.create_kernel(not_allowed)
assert False
except ValueError as e:
print(e)
```
%% Output
Field dst is written twice at the same location
%% Cell type:markdown id: tags:
This situation should be resolved by introducing temporary variables
%% Cell type:code id: tags:
``` python
tmp_var = sp.Symbol("a")
@ps.kernel
def allowed():
tmp_var @= src[0, 1] + src[1, 0]
dst[0, 0] @= 2 * tmp_var
ast = ps.create_kernel(allowed)
ps.show_code(ast)
```
%% Output
......
doc/notebooks/06_tutorial_phasefield_dentritic_growth.ipynb
View file @ 5600b6b6
Source diff could not be displayed: it is too large. Options to address this: view the blob.
doc/notebooks/demo_assignment_collection.ipynb
View file @ 5600b6b6
%% Cell type:code id: tags:
``` python
from pystencils.session import *
```
%% Cell type:markdown id: tags:
# Demo: Assignment collections and simplification
## Assignment collections
The assignment collection class helps to formulate and simplify assignments for numerical kernels.
An ``AssignmentCollection`` is an ordered collection of assignments, together with an optional ordered collection of subexpressions, that are required to evaluate the main assignments. There are various simplification rules available that operate on ``AssignmentCollection``s.
%% Cell type:markdown id: tags:
We start by defining some stencil update rule. Here we also use the *pystencils* ``Field``, note however that the assignment collection module works purely on the *sympy* level.
%% Cell type:code id: tags:
``` python
a,b,c = sp.symbols("a b c")
f = ps.fields("f(2) : [2D]")
g = ps.fields("g(2) : [2D]")
a1 = ps.Assignment(f[0,0](1), (a**2 +b) * f[0,1] + \
a1 = ps.Assignment(g[0,0](1), (a**2 +b) * f[0,1] + \
(a**2 - c) * f[1,0] + \
(a**2 - 2*c) * f[-1,0] + \
(a**2) * f[0, -1])
a2 = ps.Assignment(f[0,0](0), (c**2 +b) * f[0,1] + \
a2 = ps.Assignment(g[0,0](0), (c**2 +b) * f[0,1] + \
(c**2 - c) * f[1,0] + \
(c**2 - 2*c) * f[-1,0] + \
(c**2 - a**2) * f[0, -1])
ac = ps.AssignmentCollection([a1, a2], subexpressions=[])
ac
```
%% Output
Equation Collection for f_C^1,f_C^0
AssignmentCollection: g_C^0, g_C^1 <- f(f_N^0, b, f_S^0, f_E^0, a, f_W^0, c)
%% Cell type:markdown id: tags:
*sympy* operations can be applied on an assignment collection: In this example we first expand the collection, then look for common subexpressions.
%% Cell type:code id: tags:
``` python
expand_all = ps.simp.apply_to_all_assignments(sp.expand)
expandedEc = expand_all(ac)
```
%% Cell type:code id: tags:
``` python
ac_cse = ps.simp.sympy_cse(expandedEc)
ac_cse
```
%% Output
Equation Collection for f_C^1,f_C^0
AssignmentCollection: g_C^0, g_C^1 <- f(f_N^0, b, f_S^0, f_E^0, a, f_W^0, c)
%% Cell type:markdown id: tags:
Symbols occuring in assignment collections are classified into 3 categories:
- ``free_symbols``: symbols that occur in right-hand-sides but never on left-hand-sides
- ``bound_symbols``: symbols that occur on left-hand-sides
- ``defined_symbols``: symbols that occur on left-hand-sides of a main assignment
%% Cell type:code id: tags:
``` python
ac_cse.free_symbols
```
%% Output
$$\left\{{{f}_{E}^{0}}, {{f}_{N}^{0}}, {{f}_{S}^{0}}, {{f}_{W}^{0}}, a, b, c\right\}$$
set([f_E__0, f_N__0, f_S__0, f_W__0, a, b, c])
$\displaystyle \left\{{f}_{(1,0)}^{0}, {f}_{(0,1)}^{0}, {f}_{(0,-1)}^{0}, {f}_{(-1,0)}^{0}, a, b, c\right\}$
{f_E__0, f_N__0, f_S__0, f_W__0, a, b, c}
%% Cell type:code id: tags:
``` python
ac_cse.bound_symbols
```
%% Output
$$\left\{{{f}_{C}^{0}}, {{f}_{C}^{1}}, \xi_{0}, \xi_{1}, \xi_{2}, \xi_{3}\right\}$$
set([f_C__0, f_C__1, ξ₀, ξ₁, ξ₂, ξ₃])
$\displaystyle \left\{{g}_{(0,0)}^{0}, {g}_{(0,0)}^{1}, \xi_{0}, \xi_{1}, \xi_{2}, \xi_{3}\right\}$
{g_C__0, g_C__1, ξ₀, ξ₁, ξ₂, ξ₃}
%% Cell type:code id: tags:
``` python
ac_cse.defined_symbols
```
%% Output
$$\left\{{{f}_{C}^{0}}, {{f}_{C}^{1}}\right\}$$
set([f_C__0, f_C__1])
$\displaystyle \left\{{g}_{(0,0)}^{0}, {g}_{(0,0)}^{1}\right\}$
{g_C__0, g_C__1}
%% Cell type:markdown id: tags:
Assignment collections can be splitted up, and merged together. For splitting, a list of symbols that occur on the left-hand-side in the main assignments has to be passed. The returned assignment collection only contains these main assignments together with all necessary subexpressions.
%% Cell type:code id: tags:
``` python
ac_f0 = ac_cse.new_filtered([f(0)])
ac_f1 = ac_cse.new_filtered([f(1)])
ac_f0 = ac_cse.new_filtered([g(0)])
ac_f1 = ac_cse.new_filtered([g(1)])
ac_f1
```
%% Output
Equation Collection for f_C^1
AssignmentCollection: g_C^1, <- f(f_N^0, b, f_S^0, f_E^0, a, f_W^0, c)
%% Cell type:markdown id: tags:
Note here that $\xi_4$ is no longer part of the subexpressions, since it is not used in the main assignment of $f_C^1$.
If we merge both collections together, we end up with the original collection.
%% Cell type:code id: tags:
``` python
ac_f0.new_merged(ac_f1)
```
%% Output
Equation Collection for f_C^0,f_C^1
AssignmentCollection: g_C^0, g_C^1 <- f(f_N^0, b, f_S^0, f_E^0, a, f_W^0, c)
%% Cell type:markdown id: tags:
There is also a method that inserts all subexpressions into the main assignments. This is the inverse operation of common subexpression elimination.
%% Cell type:code id: tags:
``` python
assert sp.simplify(ac_f0.new_without_subexpressions().main_assignments[0].rhs - a2.rhs) == 0
ac_f0.new_without_subexpressions()
```
%% Output
Equation Collection for f_C^0
AssignmentCollection: g_C^0, <- f(f_N^0, b, f_S^0, f_E^0, a, f_W^0, c)
%% Cell type:markdown id: tags:
To evaluate an assignment collection, use the ``lambdify`` method. It is very similar to *sympy*s ``lambdify`` function.
%% Cell type:code id: tags:
``` python
evalFct = ac_cse.lambdify([f[0,1], f[1,0]], # new parameters of returned function
fixed_symbols={a:1, b:2, c:3, f[0,-1]: 4, f[-1,0]: 5}) # fix values of other symbols
evalFct(2,1)
```
%% Output
$$\left \{ {{f}_{C}^{0}} : 75, \quad {{f}_{C}^{1}} : -17\right \}$$
{f_C__0: 75, f_C__1: -17}
$\displaystyle \left\{ {g}_{(0,0)}^{0} : 75, \ {g}_{(0,0)}^{1} : -17\right\}$
{g_C__0: 75, g_C__1: -17}
%% Cell type:markdown id: tags:
lambdify is rather slow for evaluation. The intended way to evaluate an assignment collection is *pystencils* i.e. create a fast kernel, that applies the update at every site of a structured grid. The collection can be directly passed to the `create_kernel` function.
%% Cell type:code id: tags:
``` python
func = ps.create_kernel(ac_cse).compile()
```
%% Cell type:markdown id: tags:
## Simplification Strategies
In above examples, we already applied simplification rules to assignment collections. Simplification rules are functions that take, as a single argument, an assignment collection and return an modified/simplified copy of it. The ``SimplificationStrategy`` class holds a list of simplification rules and can apply all of them in the specified order. Additionally it provides useful printing and reporting functions.
We start by creating a simplification strategy, consisting of the expand and CSE simplifications we have already applied above:
%% Cell type:code id: tags:
``` python
strategy = ps.simp.SimplificationStrategy()
strategy.add(ps.simp.apply_to_all_assignments(sp.expand))
strategy.add(ps.simp.sympy_cse)
```
%% Cell type:markdown id: tags:
This strategy can be applied to any assignment collection:
%% Cell type:code id: tags:
``` python
strategy(ac)
```
%% Output
Equation Collection for f_C^1,f_C^0
AssignmentCollection: g_C^0, g_C^1 <- f(f_N^0, b, f_S^0, f_E^0, a, f_W^0, c)
%% Cell type:markdown id: tags:
The strategy can also print the simplification results at each stage.
The report contains information about the number of operations after each simplification as well as the runtime of each simplification routine.
%% Cell type:code id: tags:
``` python
strategy.create_simplification_report(ac)
```
%% Output
<pystencils.simp.simplificationstrategy.SimplificationStrategy.create_simplification_report.<locals>.Report at 0x7f9be404fda0>
<pystencils.simp.simplificationstrategy.SimplificationStrategy.create_simplification_report.<locals>.Report at 0x147de3e90>
%% Cell type:markdown id: tags:
The strategy can also print the full collection after each simplification...
%% Cell type:code id: tags:
``` python
strategy.show_intermediate_results(ac)
```
%% Output
<pystencils.simp.simplificationstrategy.SimplificationStrategy.show_intermediate_results.<locals>.IntermediateResults at 0x7f9bad688dd8>
<pystencils.simp.simplificationstrategy.SimplificationStrategy.show_intermediate_results.<locals>.IntermediateResults at 0x147e09c90>
%% Cell type:markdown id: tags:
... or only specific assignments for better readability
%% Cell type:code id: tags:
``` python
strategy.show_intermediate_results(ac, symbols=[f(1)])
strategy.show_intermediate_results(ac, symbols=[g(1)])
```
%% Output
<pystencils.simp.simplificationstrategy.SimplificationStrategy.show_intermediate_results.<locals>.IntermediateResults at 0x7f9bad688b00>
<pystencils.simp.simplificationstrategy.SimplificationStrategy.show_intermediate_results.<locals>.IntermediateResults at 0x1265a1b90>
%% Cell type:code id: tags:
``` python
```
......
pyproject.toml 0 → 100644
View file @ 5600b6b6
[project]
name = "pystencils"
description = "Speeding up stencil computations on CPUs and GPUs"
dynamic = ["version"]
readme = "README.md"
authors = [
{ name = "Martin Bauer" },
{ name = "Jan Hönig " },
{ name = "Markus Holzer" },
{ name = "Frederik Hennig" },
{ email = "cs10-codegen@fau.de" },
]
license = { file = "COPYING.txt" }
requires-python = ">=3.10"
dependencies = ["sympy>=1.9,<=1.12.1", "numpy>=1.8.0", "appdirs", "joblib", "pyyaml", "fasteners"]
classifiers = [
"Development Status :: 4 - Beta",
"Framework :: Jupyter",
"Topic :: Software Development :: Code Generators",
"Topic :: Scientific/Engineering :: Physics",
"Intended Audience :: Developers",
"Intended Audience :: Science/Research",
"License :: OSI Approved :: GNU Affero General Public License v3 or later (AGPLv3+)",
]
[project.urls]
"Bug Tracker" = "https://i10git.cs.fau.de/pycodegen/pystencils/-/issues"
"Documentation" = "https://pycodegen.pages.i10git.cs.fau.de/pystencils/"
"Source Code" = "https://i10git.cs.fau.de/pycodegen/pystencils"
[project.optional-dependencies]
gpu = ['cupy']
alltrafos = ['islpy', 'py-cpuinfo']
bench_db = ['blitzdb', 'pymongo', 'pandas']
interactive = [
'matplotlib',
'ipy_table',
'imageio',
'jupyter',
'pyevtk',
'rich',
'graphviz',
]
use_cython = [
'Cython'
]
doc = [
'sphinx',
'sphinx_rtd_theme',
'nbsphinx',
'sphinxcontrib-bibtex',
'sphinx_autodoc_typehints',
'pandoc',
]
tests = [
'pytest',
'pytest-cov',
'pytest-html',
'ansi2html',
'pytest-xdist',
'flake8',
'nbformat',
'nbconvert',
'ipython',
'matplotlib',
'py-cpuinfo',
'randomgen>=1.18',
]
[build-system]
requires = [
"setuptools>=61",
"versioneer[toml]>=0.29",
# 'Cython'
]
build-backend = "setuptools.build_meta"
[tool.setuptools.package-data]
pystencils = [
"include/*.h",
"boundaries/createindexlistcython.pyx"
]
[tool.setuptools.packages.find]
where = ["src"]
include = ["pystencils", "pystencils.*"]
namespaces = false
[tool.versioneer]
# See the docstring in versioneer.py for instructions. Note that you must
# re-run 'versioneer.py setup' after changing this section, and commit the
# resulting files.
VCS = "git"
style = "pep440"
versionfile_source = "src/pystencils/_version.py"
versionfile_build = "pystencils/_version.py"
tag_prefix = "release/"
parentdir_prefix = "pystencils-"
pystencils/include/PyStencilsField.h deleted 100644 → 0
View file @ 5a059a11
#pragma once
extern "C++" {
#ifdef __CUDA_ARCH__
template <typename DTYPE_T, std::size_t DIMENSION> struct PyStencilsField {
DTYPE_T *data;
DTYPE_T shape[DIMENSION];
DTYPE_T stride[DIMENSION];
};
#else
#include <array>
template <typename DTYPE_T, std::size_t DIMENSION> struct PyStencilsField {
DTYPE_T *data;
std::array<DTYPE_T, DIMENSION> shape;
std::array<DTYPE_T, DIMENSION> stride;
};
#endif
}
pystencils/include/gpu_defines.h deleted 100644 → 0
View file @ 5a059a11
#pragma once
#define POS_INFINITY __int_as_float(0x7f800000)
#define INFINITY POS_INFINITY
#define NEG_INFINITY __int_as_float(0xff800000)
#ifdef __HIPCC_RTC__
typedef __hip_uint8_t uint8_t;
typedef __hip_int8_t int8_t;
typedef __hip_uint16_t uint16_t;
typedef __hip_int16_t int16_t;
#endif
pystencils/node_collection.py deleted 100644 → 0
View file @ 5a059a11
from typing import List, Union
import sympy
import sympy as sp
from sympy.codegen import Assignment
from sympy.codegen.rewriting import ReplaceOptim, optimize
from pystencils.astnodes import Block, Node, SympyAssignment
from pystencils.backends.cbackend import CustomCodeNode
from pystencils.functions import DivFunc
from pystencils.simp import AssignmentCollection
class NodeCollection:
def __init__(self, assignments: List[Union[Node, Assignment]]):
self.all_assignments = assignments
if all((isinstance(a, Assignment) for a in assignments)):
self.is_Nodes = False
self.is_Assignments = True
elif all((isinstance(n, Node) for n in assignments)):
self.is_Nodes = True
self.is_Assignments = False
else:
raise ValueError(f'The list "{assignments}" is mixed. Pass either a list of "pystencils.Assignments" '
f'or a list of "pystencils.astnodes.Node')
self.simplification_hints = {}
@staticmethod
def from_assignment_collection(assignment_collection: AssignmentCollection):
nodes = list()
for assignemt in assignment_collection.all_assignments:
if isinstance(assignemt, Assignment):
nodes.append(SympyAssignment(assignemt.lhs, assignemt.rhs))
elif isinstance(assignemt, Node):
nodes.append(assignemt)
else:
raise ValueError(f"Unknown node in the AssignmentCollection: {assignemt}")
return NodeCollection(nodes)
def evaluate_terms(self):
evaluate_constant_terms = ReplaceOptim(
lambda e: hasattr(e, 'is_constant') and e.is_constant and not e.is_integer,
lambda p: p.evalf()
)
evaluate_pow = ReplaceOptim(
lambda e: e.is_Pow and e.exp.is_Integer and abs(e.exp) <= 8,
lambda p: sp.UnevaluatedExpr(sp.Mul(*([p.base] * +p.exp), evaluate=False)) if p.exp > 0 else
(DivFunc(sp.Integer(1), p.base) if p.exp == -1 else
DivFunc(sp.Integer(1), sp.UnevaluatedExpr(sp.Mul(*([p.base] * -p.exp), evaluate=False))))
)
sympy_optimisations = [evaluate_constant_terms, evaluate_pow]
if self.is_Nodes:
def visitor(node):
if isinstance(node, CustomCodeNode):
return node
elif isinstance(node, Block):
return node.func([visitor(child) for child in node.args])
elif isinstance(node, Node):
return node.func(*[visitor(child) for child in node.args])
elif isinstance(node, sympy.Basic):
return optimize(node, sympy_optimisations)
else:
raise NotImplementedError(f'{node} {type(node)} has no valid visitor')
self.all_assignments = [visitor(assignment) for assignment in self.all_assignments]
else:
self.all_assignments = [Assignment(a.lhs, optimize(a.rhs, sympy_optimisations))
if hasattr(a, 'lhs')
else a for a in self.all_assignments]
pystencils_tests/test_indexed_kernels.py deleted 100644 → 0
View file @ 5a059a11
import numpy as np
import pytest
import pystencils as ps
from pystencils import Assignment, Field, CreateKernelConfig, create_kernel, Target
def test_indexed_kernel():
arr = np.zeros((3, 4))
dtype = np.dtype([('x', int), ('y', int), ('value', arr.dtype)])
index_arr = np.zeros((3,), dtype=dtype)
index_arr[0] = (0, 2, 3.0)
index_arr[1] = (1, 3, 42.0)
index_arr[2] = (2, 1, 5.0)
indexed_field = Field.create_from_numpy_array('index', index_arr)
normal_field = Field.create_from_numpy_array('f', arr)
update_rule = Assignment(normal_field[0, 0], indexed_field('value'))
config = CreateKernelConfig(index_fields=[indexed_field])
ast = create_kernel([update_rule], config=config)
kernel = ast.compile()
kernel(f=arr, index=index_arr)
code = ps.get_code_str(kernel)
for i in range(index_arr.shape[0]):
np.testing.assert_allclose(arr[index_arr[i]['x'], index_arr[i]['y']], index_arr[i]['value'], atol=1e-13)
def test_indexed_gpu_kernel():
pytest.importorskip("cupy")
import cupy as cp
arr = np.zeros((3, 4))
dtype = np.dtype([('x', int), ('y', int), ('value', arr.dtype)])
index_arr = np.zeros((3,), dtype=dtype)
index_arr[0] = (0, 2, 3.0)
index_arr[1] = (1, 3, 42.0)
index_arr[2] = (2, 1, 5.0)
indexed_field = Field.create_from_numpy_array('index', index_arr)
normal_field = Field.create_from_numpy_array('f', arr)
update_rule = Assignment(normal_field[0, 0], indexed_field('value'))
config = CreateKernelConfig(target=Target.GPU, index_fields=[indexed_field])
ast = create_kernel([update_rule], config=config)
kernel = ast.compile()
gpu_arr = cp.asarray(arr)
gpu_index_arr = cp.ndarray(index_arr.shape, dtype=index_arr.dtype)
gpu_index_arr.set(index_arr)
kernel(f=gpu_arr, index=gpu_index_arr)
arr = gpu_arr.get()
for i in range(index_arr.shape[0]):
np.testing.assert_allclose(arr[index_arr[i]['x'], index_arr[i]['y']], index_arr[i]['value'], atol=1e-13)
pystencils_tests/test_jupyter_extensions.ipynb deleted 100644 → 0
View file @ 5a059a11
%% Cell type:code id: tags:
``` python
from pystencils.session import *
```
%% Cell type:code id: tags:
``` python
dh = ps.create_data_handling(domain_size=(256, 256), periodicity=True)
c_field = dh.add_array('c')
dh.fill("c", 0.0, ghost_layers=True)
```
%% Cell type:code id: tags:
``` python
for x in range(129):
for y in range(258):
dh.cpu_arrays['c'][x, y] = 1.0
```
%% Cell type:code id: tags:
``` python
plt.scalar_field(dh.cpu_arrays["c"])
```
%% Output
<matplotlib.image.AxesImage at 0x7fcb7d253710>
%% Cell type:code id: tags:
``` python
ur = ps.Assignment(c_field[0, 0], c_field[1, 0])
ast = ps.create_kernel(ur, target=dh.default_target, cpu_openmp=True)
kernel = ast.compile()
```
%% Cell type:code id: tags:
``` python
c_sync = dh.synchronization_function_cpu(['c'])
```
%% Cell type:code id: tags:
``` python
def timeloop(steps=10):
for i in range(steps):
c_sync()
dh.run_kernel(kernel)
return dh.gather_array('c')
```
%% Cell type:code id: tags:
``` python
ps.jupyter.set_display_mode('video')
```
%% Cell type:code id: tags:
``` python
ani = ps.plot.scalar_field_animation(timeloop, rescale=True, frames=12)
ps.jupyter.display_animation(ani)
```
%% Output
<IPython.core.display.HTML object>
%% Cell type:code id: tags:
``` python
ps.jupyter.set_display_mode('image_update')
```
%% Cell type:code id: tags:
``` python
ani = ps.plot.scalar_field_animation(timeloop, rescale=True, frames=12)
ps.jupyter.display_animation(ani)
```
%% Output
%% Cell type:code id: tags:
``` python
def grid_update_function(image):
for i in range(40):
c_sync()
dh.run_kernel(kernel)
return dh.gather_array('c')
```
%% Cell type:code id: tags:
``` python
animation = ps.jupyter.make_imshow_animation(dh.cpu_arrays["c"], grid_update_function, frames=300)
```
%% Output
%% Cell type:code id: tags:
``` python
ps.jupyter.set_display_mode("video")
ps.jupyter.set_display_mode("window")
ps.jupyter.set_display_mode("image_update")
ps.jupyter.activate_ipython()
```
pystencils_tests/test_print_infinity.py deleted 100644 → 0
View file @ 5a059a11
import pytest
import pystencils
from sympy import oo
@pytest.mark.parametrize('type', ('float32', 'float64', 'int64'))
@pytest.mark.parametrize('negative', (False, 'Negative'))
@pytest.mark.parametrize('target', (pystencils.Target.CPU, pystencils.Target.GPU))
def test_print_infinity(type, negative, target):
x = pystencils.fields(f'x: {type}[1d]')
if negative:
assignment = pystencils.Assignment(x.center, -oo)
else:
assignment = pystencils.Assignment(x.center, oo)
ast = pystencils.create_kernel(assignment, data_type=type, target=target)
if target == pystencils.Target.GPU:
pytest.importorskip('cupy')
ast.compile()
print(ast.compile().code)
pystencils_tests/test_print_unsupported_node.py deleted 100644 → 0
View file @ 5a059a11
# -*- coding: utf-8 -*-
#
# Copyright © 2019 Stephan Seitz <stephan.seitz@fau.de>
#
# Distributed under terms of the GPLv3 license.
"""
"""
import pytest
import pystencils
from pystencils.backends.cbackend import CBackend
class UnsupportedNode(pystencils.astnodes.Node):
def __init__(self):
super().__init__()
def test_print_unsupported_node():
with pytest.raises(NotImplementedError, match='CBackend does not support node of type UnsupportedNode'):
CBackend()(UnsupportedNode())
pystencils_tests/test_sliced_iteration.py deleted 100644 → 0
View file @ 5a059a11
import numpy as np
import sympy as sp
from pystencils import Assignment, Field, TypedSymbol, create_kernel, make_slice
from pystencils.simp import sympy_cse_on_assignment_list
def test_sliced_iteration():
size = (4, 4)
src_arr = np.ones(size)
dst_arr = np.zeros_like(src_arr)
src_field = Field.create_from_numpy_array('src', src_arr)
dst_field = Field.create_from_numpy_array('dst', dst_arr)
a, b = sp.symbols("a b")
update_rule = Assignment(dst_field[0, 0],
(a * src_field[0, 1] + a * src_field[0, -1] +
b * src_field[1, 0] + b * src_field[-1, 0]) / 4)
x_end = TypedSymbol("x_end", "int")
s = make_slice[1:x_end, 1]
x_end_value = size[1] - 1
kernel = create_kernel(sympy_cse_on_assignment_list([update_rule]), iteration_slice=s).compile()
kernel(src=src_arr, dst=dst_arr, a=1.0, b=1.0, x_end=x_end_value)
expected_result = np.zeros(size)
expected_result[1:x_end_value, 1] = 1
np.testing.assert_almost_equal(expected_result, dst_arr)
pytest.ini
View file @ 5600b6b6
[pytest]
testpaths = src tests doc/notebooks
pythonpath = src
python_files = test_*.py *_test.py scenario_*.py
norecursedirs = *.egg-info .git .cache .ipynb_checkpoints htmlcov
addopts = --doctest-modules --durations=20 --cov-config pytest.ini
......@@ -17,20 +19,21 @@ filterwarnings =
[run]
branch = True
source = pystencils
pystencils_tests
source = src/pystencils
tests
omit = doc/*
pystencils_tests/*
tests/*
setup.py
quicktest.py
conftest.py
versioneer.py
pystencils/jupytersetup.py
pystencils/cpu/msvc_detection.py
pystencils/sympy_gmpy_bug_workaround.py
pystencils/cache.py
pystencils/pacxx/benchmark.py
pystencils/_version.py
src/pystencils/jupytersetup.py
src/pystencils/cpu/msvc_detection.py
src/pystencils/sympy_gmpy_bug_workaround.py
src/pystencils/cache.py
src/pystencils/pacxx/benchmark.py
src/pystencils/_version.py
venv/
[report]
......