| Name | gt4py JSON |
| Version |
1.0.4
JSON |
| download |
| home_page | None |
| Summary | Python library for generating high-performance implementations of stencil kernels for weather and climate modeling from a domain-specific language (DSL) |
| upload_time | 2024-09-23 09:04:12 |
| maintainer | None |
| docs_url | None |
| author | None |
| requires_python | >=3.8 |
| license | None |
| keywords |
gridtools
stencil
weather
climate
performance
portable
hpc
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
[](https://GridTools.github.io/gt4py)
[](https://join.slack.com/t/gridtools/shared_invite/zt-1mceuj747-59swuowC3MKAuCFyNAnc1g)
[](https://github.com/GridTools/gt4py/actions/workflows/daily-ci.yml)
# GT4Py: GridTools for Python
GT4Py is a Python library for generating high performance implementations of stencil kernels from a high-level definition using regular Python functions. GT4Py is part of the GridTools framework, a set of libraries and utilities to develop performance portable applications in the area of weather and climate modeling.
**NOTE:** The `gt4py.next` subpackage contains a new version of GT4Py which is not compatible with the current _stable_ version defined in `gt4py.cartesian`. The new version is highly experimental, it only works with unstructured meshes and it requires `python >= 3.10`.
## 📃 Description
GT4Py is a Python library for expressing computational motifs as found in weather and climate applications. These computations are expressed in a domain specific language (GTScript) which is translated to high-performance implementations for CPUs and GPUs.
The DSL expresses computations on a 3-dimensional Cartesian grid. The horizontal axes (`I`, `J`) are always computed in parallel, while the vertical (`K`) can be iterated in sequential, forward or backward, order. Cartesian offsets are expressed relative to a center index.
In addition, GT4Py provides functions to allocate arrays with memory layout suited for a particular backend.
The following backends are supported:
- `numpy`: Pure-Python backend
- `gt:cpu_ifirst`: GridTools C++ CPU backend using `I`-first data ordering
- `gt:cpu_kfirst`: GridTools C++ CPU backend using `K`-first data ordering
- `gt:gpu`: GridTools backend for CUDA
- `cuda`: CUDA backend minimally using utilities from GridTools
- `dace:cpu`: Dace code-generated CPU backend
- `dace:gpu`: Dace code-generated GPU backend
## 🚜 Installation
GT4Py can be installed as a regular Python package using `pip` (or any other PEP-517 frontend). As usual, we strongly recommended to create a new virtual environment to work on this project.
The performance backends also require the `Boost <https://www.boost.org/>`\_\_ library, a dependency of [GridTools C++](https://github.com/GridTools/gridtools), which needs to be installed by the user.
## ⚙ Configuration
If GridTools or Boost are not found in the compiler's standard include path, or a custom version is desired, then a couple configuration environment variables will allow the compiler to use them:
- `GT_INCLUDE_PATH`: Path to the GridTools installation.
- `BOOST_ROOT`: Path to a boost installation.
Other commonly used environment variables are:
- `CUDA_ARCH`: Set the compute capability of the NVIDIA GPU if it is not detected automatically by `cupy`.
- `CXX`: Set the C++ compiler.
- `GT_CACHE_DIR_NAME`: Name of the compiler's cache directory (defaults to `.gt_cache`)
- `GT_CACHE_ROOT`: Path to the compiler cache (defaults to `./`)
More options and details are available in [`config.py`](https://github.com/GridTools/gt4py/blob/main/src/gt4py/cartesian/config.py>).
## 📖 Documentation
GT4Py uses Sphinx documentation. To build the documentation install the dependencies in `requirements-dev.txt`
```bash
pip install -r ./gt4py/requirements-dev.txt
```
and then build the docs with
```bash
cd gt4py/docs/user/cartesian
make html # run 'make help' for a list of targets
```
## 🛠 Development Instructions
Follow the installation instructions below to initialize a development virtual environment containing an _editable_ installation of the GT4Py package. Make sure you read the [CONTRIBUTING.md](CONTRIBUTING.md) and [CODING_GUIDELINES.md](CODING_GUIDELINES.md) documents before you start working on the project.
### Recommended Installation using `tox`
If [tox](https://tox.wiki/en/latest/) is already installed in your system (`tox` is available in PyPI and many other package managers), the easiest way to create a virtual environment ready for development is:
```bash
# Clone the repository
git clone https://github.com/gridtools/gt4py.git
cd gt4py
# Create the development environment in any location (usually `.venv`)
# selecting one of the following templates:
# dev-py310 -> base environment
# dev-py310-atlas -> base environment + atlas4py bindings
tox devenv -e dev-py310 .venv
# Finally, activate the environment
source .venv/bin/activate
```
### Manual Installation
Alternatively, a development environment can be created from scratch installing the frozen dependencies packages :
```bash
# Clone the repository
git clone https://github.com/gridtools/gt4py.git
cd gt4py
# Create a (Python 3.10) virtual environment (usually at `.venv`)
python3.10 -m venv .venv
# Activate the virtual environment and update basic packages
source .venv/bin/activate
pip install --upgrade wheel setuptools pip
# Install the required development tools
pip install -r requirements-dev.txt
# Install GT4Py project in editable mode
pip install -e .
# Optionally, install atlas4py bindings directly from the repo
# pip install git+https://github.com/GridTools/atlas4py#egg=atlas4py
```
## ⚖️ License
GT4Py is licensed under the terms of the [BSD-3-Clause](https://github.com/GridTools/gt4py/blob/main/LICENSE.txt).
Raw data
{
"_id": null,
"home_page": null,
"name": "gt4py",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.8",
"maintainer_email": null,
"keywords": "gridtools, stencil, weather, climate, performance, portable, hpc",
"author": null,
"author_email": "ETH Zurich <gridtools@cscs.ch>",
"download_url": "https://files.pythonhosted.org/packages/b4/66/cd8616b7e77f8e4d3680de77111102b8a82cbcdbab1c6a99ac6bb4126be2/gt4py-1.0.4.tar.gz",
"platform": "Linux",
"description": "[](https://GridTools.github.io/gt4py)\n\n\n[](https://join.slack.com/t/gridtools/shared_invite/zt-1mceuj747-59swuowC3MKAuCFyNAnc1g)\n\n[](https://github.com/GridTools/gt4py/actions/workflows/daily-ci.yml)\n\n\n\n\n\n\n# GT4Py: GridTools for Python\n\nGT4Py is a Python library for generating high performance implementations of stencil kernels from a high-level definition using regular Python functions. GT4Py is part of the GridTools framework, a set of libraries and utilities to develop performance portable applications in the area of weather and climate modeling.\n\n**NOTE:** The `gt4py.next` subpackage contains a new version of GT4Py which is not compatible with the current _stable_ version defined in `gt4py.cartesian`. The new version is highly experimental, it only works with unstructured meshes and it requires `python >= 3.10`.\n\n## \ud83d\udcc3 Description\n\nGT4Py is a Python library for expressing computational motifs as found in weather and climate applications. These computations are expressed in a domain specific language (GTScript) which is translated to high-performance implementations for CPUs and GPUs.\n\nThe DSL expresses computations on a 3-dimensional Cartesian grid. The horizontal axes (`I`, `J`) are always computed in parallel, while the vertical (`K`) can be iterated in sequential, forward or backward, order. Cartesian offsets are expressed relative to a center index.\n\nIn addition, GT4Py provides functions to allocate arrays with memory layout suited for a particular backend.\n\nThe following backends are supported:\n\n- `numpy`: Pure-Python backend\n- `gt:cpu_ifirst`: GridTools C++ CPU backend using `I`-first data ordering\n- `gt:cpu_kfirst`: GridTools C++ CPU backend using `K`-first data ordering\n- `gt:gpu`: GridTools backend for CUDA\n- `cuda`: CUDA backend minimally using utilities from GridTools\n- `dace:cpu`: Dace code-generated CPU backend\n- `dace:gpu`: Dace code-generated GPU backend\n\n## \ud83d\ude9c Installation\n\nGT4Py can be installed as a regular Python package using `pip` (or any other PEP-517 frontend). As usual, we strongly recommended to create a new virtual environment to work on this project.\n\nThe performance backends also require the `Boost <https://www.boost.org/>`\\_\\_ library, a dependency of [GridTools C++](https://github.com/GridTools/gridtools), which needs to be installed by the user.\n\n## \u2699 Configuration\n\nIf GridTools or Boost are not found in the compiler's standard include path, or a custom version is desired, then a couple configuration environment variables will allow the compiler to use them:\n\n- `GT_INCLUDE_PATH`: Path to the GridTools installation.\n- `BOOST_ROOT`: Path to a boost installation.\n\nOther commonly used environment variables are:\n\n- `CUDA_ARCH`: Set the compute capability of the NVIDIA GPU if it is not detected automatically by `cupy`.\n- `CXX`: Set the C++ compiler.\n- `GT_CACHE_DIR_NAME`: Name of the compiler's cache directory (defaults to `.gt_cache`)\n- `GT_CACHE_ROOT`: Path to the compiler cache (defaults to `./`)\n\nMore options and details are available in [`config.py`](https://github.com/GridTools/gt4py/blob/main/src/gt4py/cartesian/config.py>).\n\n## \ud83d\udcd6 Documentation\n\nGT4Py uses Sphinx documentation. To build the documentation install the dependencies in `requirements-dev.txt`\n\n```bash\npip install -r ./gt4py/requirements-dev.txt\n```\n\nand then build the docs with\n\n```bash\ncd gt4py/docs/user/cartesian\nmake html # run 'make help' for a list of targets\n```\n\n## \ud83d\udee0 Development Instructions\n\nFollow the installation instructions below to initialize a development virtual environment containing an _editable_ installation of the GT4Py package. Make sure you read the [CONTRIBUTING.md](CONTRIBUTING.md) and [CODING_GUIDELINES.md](CODING_GUIDELINES.md) documents before you start working on the project.\n\n### Recommended Installation using `tox`\n\nIf [tox](https://tox.wiki/en/latest/) is already installed in your system (`tox` is available in PyPI and many other package managers), the easiest way to create a virtual environment ready for development is:\n\n```bash\n# Clone the repository\ngit clone https://github.com/gridtools/gt4py.git\ncd gt4py\n\n# Create the development environment in any location (usually `.venv`)\n# selecting one of the following templates:\n# dev-py310 -> base environment\n# dev-py310-atlas -> base environment + atlas4py bindings\ntox devenv -e dev-py310 .venv\n\n# Finally, activate the environment\nsource .venv/bin/activate\n```\n\n### Manual Installation\n\nAlternatively, a development environment can be created from scratch installing the frozen dependencies packages :\n\n```bash\n# Clone the repository\ngit clone https://github.com/gridtools/gt4py.git\ncd gt4py\n\n# Create a (Python 3.10) virtual environment (usually at `.venv`)\npython3.10 -m venv .venv\n\n# Activate the virtual environment and update basic packages\nsource .venv/bin/activate\npip install --upgrade wheel setuptools pip\n\n# Install the required development tools\npip install -r requirements-dev.txt\n# Install GT4Py project in editable mode\npip install -e .\n\n# Optionally, install atlas4py bindings directly from the repo\n# pip install git+https://github.com/GridTools/atlas4py#egg=atlas4py\n```\n\n## \u2696\ufe0f License\n\nGT4Py is licensed under the terms of the [BSD-3-Clause](https://github.com/GridTools/gt4py/blob/main/LICENSE.txt).\n",
"bugtrack_url": null,
"license": null,
"summary": "Python library for generating high-performance implementations of stencil kernels for weather and climate modeling from a domain-specific language (DSL)",
"version": "1.0.4",
"project_urls": {
"Documentation": "https://gridtools.github.io/gt4py",
"Homepage": "https://gridtools.github.io/",
"Source": "https://github.com/GridTools/gt4py"
},
"split_keywords": [
"gridtools",
" stencil",
" weather",
" climate",
" performance",
" portable",
" hpc"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "63a18693031598dcedcfbf54804932d2d0584d2782dda5c788a3516c7b77d973",
"md5": "01f9d67e4d45165e45b1198aeea914db",
"sha256": "ffb358856b2996757b4891cb969ef4db530d682dcbe7ac5b7193278a01308a0d"
},
"downloads": -1,
"filename": "gt4py-1.0.4-py3-none-any.whl",
"has_sig": false,
"md5_digest": "01f9d67e4d45165e45b1198aeea914db",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8",
"size": 761022,
"upload_time": "2024-09-23T09:04:10",
"upload_time_iso_8601": "2024-09-23T09:04:10.094607Z",
"url": "https://files.pythonhosted.org/packages/63/a1/8693031598dcedcfbf54804932d2d0584d2782dda5c788a3516c7b77d973/gt4py-1.0.4-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "b466cd8616b7e77f8e4d3680de77111102b8a82cbcdbab1c6a99ac6bb4126be2",
"md5": "19d5e1974a52e5c875bf2d6b11a3437d",
"sha256": "93aee19abf35c7c589fe7a78e7a7a54ce576dc1a0dce68b78edadaeb14ce76da"
},
"downloads": -1,
"filename": "gt4py-1.0.4.tar.gz",
"has_sig": false,
"md5_digest": "19d5e1974a52e5c875bf2d6b11a3437d",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8",
"size": 582955,
"upload_time": "2024-09-23T09:04:12",
"upload_time_iso_8601": "2024-09-23T09:04:12.069947Z",
"url": "https://files.pythonhosted.org/packages/b4/66/cd8616b7e77f8e4d3680de77111102b8a82cbcdbab1c6a99ac6bb4126be2/gt4py-1.0.4.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-09-23 09:04:12",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "GridTools",
"github_project": "gt4py",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"tox": true,
"lcname": "gt4py"
}
