| Name | jaxgl JSON |
| Version |
1.0.1
JSON |
| download |
| home_page | None |
| Summary | An simple JAX graphics library |
| upload_time | 2024-11-03 22:07:24 |
| maintainer | None |
| docs_url | None |
| author | None |
| requires_python | >=3.10 |
| license | None |
| keywords |
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# JaxGL
<p align="center">
<a href= "https://pypi.org/project/jaxgl/">
<img src="https://img.shields.io/badge/python-3.10%20%7C%203.11%20%7C%203.12-blue" /></a>
<a href= "https://pypi.org/project/jaxgl/">
<img src="https://img.shields.io/badge/pypi-1.0.1-green" /></a>
<a href= "https://github.com/MichaelTMatthews/Craftax/blob/main/LICENSE">
<img src="https://img.shields.io/badge/License-MIT-yellow" /></a>
<a href= "https://github.com/psf/black">
<img src="https://img.shields.io/badge/code%20style-black-000000.svg" /></a>
</p>
JaxGL is a simple and flexible graphics library written entirely in <a href="https://github.com/google/jax">JAX</a>. JaxGL was created by [Michael Matthews](https://github.com/MichaelTMatthews) and [Michael Beukman](https://github.com/Michael-Beukman) for the [Kinetix](https://github.com/FLAIROx/Kinetix) project.
# 💻 Basic Usage
```python
# 512x512 pixels
screen_size = (512, 512)
# Clear a fresh screen with a black background
clear_colour = jnp.array([0.0, 0.0, 0.0])
pixels = clear_screen(screen_size, clear_colour)
# We render to a 256x256 'patch'
patch_size = (256, 256)
triangle_renderer = make_renderer(screen_size, fragment_shader_triangle, patch_size)
# Patch position (top left corner)
pos = jnp.array([128, 128])
triangle_data = (
# Vertices (note these must be anti-clockwise)
jnp.array([[150, 200], [150, 300], [300, 150]]),
# Colour
jnp.array([255.0, 0.0, 0.0]),
)
# Render the triangle to the screen
pixels = triangle_renderer(pixels, pos, triangle_data)
```
This produces the following image:
<p align="center">
<img width="20%" src="images/simple_render.png" />
</p>
# 👨💻 Custom Shaders
Arbitrary rendering effects can be achieved by writing your own shaders.
```python
screen_size = (512, 512)
clear_colour = jnp.array([0.0, 0.0, 0.0])
pixels = clear_screen(screen_size, clear_colour)
patch_size = (256, 256)
# We make our own variation of the circle shader
# We give both a central and edge colour and interpolate between these
# Each fragment shader has access to
# position: global position in screen space
# current_frag: the current colour of the fragment (useful for transparency)
# unit_position: the position inside the patch (scaled to between 0 and 1)
# uniform: anything you want for your shader. These are the same for every fragment.
def my_shader(position, current_frag, unit_position, uniform):
centre, radius, colour_centre, colour_outer = uniform
dist = jnp.sqrt(jnp.square(position - centre).sum())
colour_interp = dist / radius
colour = colour_interp * colour_outer + (1 - colour_interp) * colour_centre
return jax.lax.select(dist < radius, colour, current_frag)
circle_renderer = make_renderer(screen_size, my_shader, patch_size)
# Patch position (top left corner)
pos = jnp.array([128, 128])
# This is the uniform that is passed to the shader
circle_data = (
# Centre
jnp.array([256.0, 256.0]),
# Radius
100.0,
# Colour centre
jnp.array([255.0, 0.0, 0.0]),
# Colour outer
jnp.array([0.0, 255.0, 0.0]),
)
# Render the triangle to the screen
pixels = circle_renderer(pixels, pos, circle_data)
```
<p align="center">
<img width="20%" src="images/custom_shader.png" />
</p>
# 🔄 In Kinetix
JaxGL is used for rendering in [Kinetix](https://github.com/FLAIROx/Kinetix). Shown below is an example robotics grasping task.
<p align="center">
<img width="40%" src="images/kinetix.png" />
</p>
# ⬇️ Installation
To use JaxGL in your work you can install via PyPi:
```commandline
pip install jaxgl
```
If you want to extend JaxGL you can install as follows:
```commandline
git clone https://github.com/FLAIROx/JaxGL
cd JaxGL
pip install -e ".[dev]"
pre-commit install
```
# 🔍 See Also
- [JAX Renderer](https://github.com/JoeyTeng/jaxrenderer) A more complete JAX renderer more suitable for 3D rendering.
- [Jax2D](https://github.com/MichaelTMatthews/Jax2D) 2D physics engine in JAX.
- [Kinetix](https://github.com/FLAIROx/Kinetix) physics-based reinforcement learning in JAX.
Raw data
{
"_id": null,
"home_page": null,
"name": "jaxgl",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.10",
"maintainer_email": null,
"keywords": null,
"author": null,
"author_email": "Michael Matthews <michael.matthews@eng.ox.ac.uk>, Michael Beukman <michael.beukman@eng.ox.ac.uk>",
"download_url": "https://files.pythonhosted.org/packages/50/4d/bfa0f8bf4fd3e85312a77cacba2a507d22feb211e9151bc4ea5da14e6050/jaxgl-1.0.1.tar.gz",
"platform": null,
"description": "# JaxGL\n<p align=\"center\">\n <a href= \"https://pypi.org/project/jaxgl/\">\n <img src=\"https://img.shields.io/badge/python-3.10%20%7C%203.11%20%7C%203.12-blue\" /></a>\n <a href= \"https://pypi.org/project/jaxgl/\">\n <img src=\"https://img.shields.io/badge/pypi-1.0.1-green\" /></a>\n <a href= \"https://github.com/MichaelTMatthews/Craftax/blob/main/LICENSE\">\n <img src=\"https://img.shields.io/badge/License-MIT-yellow\" /></a>\n <a href= \"https://github.com/psf/black\">\n <img src=\"https://img.shields.io/badge/code%20style-black-000000.svg\" /></a>\n</p>\n\nJaxGL is a simple and flexible graphics library written entirely in <a href=\"https://github.com/google/jax\">JAX</a>. JaxGL was created by [Michael Matthews](https://github.com/MichaelTMatthews) and [Michael Beukman](https://github.com/Michael-Beukman) for the [Kinetix](https://github.com/FLAIROx/Kinetix) project.\n\n# \ud83d\udcbb Basic Usage\n```python\n# 512x512 pixels\nscreen_size = (512, 512)\n\n# Clear a fresh screen with a black background\nclear_colour = jnp.array([0.0, 0.0, 0.0])\npixels = clear_screen(screen_size, clear_colour)\n\n# We render to a 256x256 'patch'\npatch_size = (256, 256)\ntriangle_renderer = make_renderer(screen_size, fragment_shader_triangle, patch_size)\n\n# Patch position (top left corner)\npos = jnp.array([128, 128])\n\ntriangle_data = (\n # Vertices (note these must be anti-clockwise)\n jnp.array([[150, 200], [150, 300], [300, 150]]),\n # Colour\n jnp.array([255.0, 0.0, 0.0]),\n)\n\n# Render the triangle to the screen\npixels = triangle_renderer(pixels, pos, triangle_data)\n```\n\nThis produces the following image:\n\n<p align=\"center\">\n <img width=\"20%\" src=\"images/simple_render.png\" />\n</p>\n\n# \ud83d\udc68\u200d\ud83d\udcbb Custom Shaders\nArbitrary rendering effects can be achieved by writing your own shaders.\n```python\nscreen_size = (512, 512)\n\nclear_colour = jnp.array([0.0, 0.0, 0.0])\npixels = clear_screen(screen_size, clear_colour)\n\npatch_size = (256, 256)\n\n# We make our own variation of the circle shader\n# We give both a central and edge colour and interpolate between these\n\n# Each fragment shader has access to\n# position: global position in screen space\n# current_frag: the current colour of the fragment (useful for transparency)\n# unit_position: the position inside the patch (scaled to between 0 and 1)\n# uniform: anything you want for your shader. These are the same for every fragment.\n\ndef my_shader(position, current_frag, unit_position, uniform):\n centre, radius, colour_centre, colour_outer = uniform\n\n dist = jnp.sqrt(jnp.square(position - centre).sum())\n colour_interp = dist / radius\n\n colour = colour_interp * colour_outer + (1 - colour_interp) * colour_centre\n\n return jax.lax.select(dist < radius, colour, current_frag)\n\ncircle_renderer = make_renderer(screen_size, my_shader, patch_size)\n\n# Patch position (top left corner)\npos = jnp.array([128, 128])\n\n# This is the uniform that is passed to the shader\ncircle_data = (\n # Centre\n jnp.array([256.0, 256.0]),\n # Radius\n 100.0,\n # Colour centre\n jnp.array([255.0, 0.0, 0.0]),\n # Colour outer\n jnp.array([0.0, 255.0, 0.0]),\n)\n\n# Render the triangle to the screen\npixels = circle_renderer(pixels, pos, circle_data)\n```\n\n<p align=\"center\">\n <img width=\"20%\" src=\"images/custom_shader.png\" />\n</p>\n\n# \ud83d\udd04 In Kinetix\nJaxGL is used for rendering in [Kinetix](https://github.com/FLAIROx/Kinetix). Shown below is an example robotics grasping task.\n<p align=\"center\">\n <img width=\"40%\" src=\"images/kinetix.png\" />\n</p>\n\n# \u2b07\ufe0f Installation\nTo use JaxGL in your work you can install via PyPi:\n```commandline\npip install jaxgl\n```\n\nIf you want to extend JaxGL you can install as follows:\n```commandline\ngit clone https://github.com/FLAIROx/JaxGL\ncd JaxGL\npip install -e \".[dev]\"\npre-commit install\n```\n\n# \ud83d\udd0d See Also\n- [JAX Renderer](https://github.com/JoeyTeng/jaxrenderer) A more complete JAX renderer more suitable for 3D rendering.\n- [Jax2D](https://github.com/MichaelTMatthews/Jax2D) 2D physics engine in JAX.\n- [Kinetix](https://github.com/FLAIROx/Kinetix) physics-based reinforcement learning in JAX.\n",
"bugtrack_url": null,
"license": null,
"summary": "An simple JAX graphics library",
"version": "1.0.1",
"project_urls": {
"Homepage": "https://github.com/FLAIROx/JaxGL",
"Issues": "https://github.com/FLAIROx/JaxGL"
},
"split_keywords": [],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "f3e13ac61628a0cc048a83a088a67603486d7fbfd68ee0e8f8a22b41a0ddecf6",
"md5": "7565e739540785cd6ac841933df6d579",
"sha256": "b9940ef8a578e4571a4a8ffcdb1e1a2b686df7f1cd46cb6e680437edbffe947e"
},
"downloads": -1,
"filename": "jaxgl-1.0.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "7565e739540785cd6ac841933df6d579",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.10",
"size": 6753,
"upload_time": "2024-11-03T22:07:23",
"upload_time_iso_8601": "2024-11-03T22:07:23.132988Z",
"url": "https://files.pythonhosted.org/packages/f3/e1/3ac61628a0cc048a83a088a67603486d7fbfd68ee0e8f8a22b41a0ddecf6/jaxgl-1.0.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "504dbfa0f8bf4fd3e85312a77cacba2a507d22feb211e9151bc4ea5da14e6050",
"md5": "2cfe88cc8463ec2d5c25f38338cea6b6",
"sha256": "051a1975f2e1932c64b98e9770473ed5c102387f69fa2edded214439de8688e0"
},
"downloads": -1,
"filename": "jaxgl-1.0.1.tar.gz",
"has_sig": false,
"md5_digest": "2cfe88cc8463ec2d5c25f38338cea6b6",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.10",
"size": 6478,
"upload_time": "2024-11-03T22:07:24",
"upload_time_iso_8601": "2024-11-03T22:07:24.623520Z",
"url": "https://files.pythonhosted.org/packages/50/4d/bfa0f8bf4fd3e85312a77cacba2a507d22feb211e9151bc4ea5da14e6050/jaxgl-1.0.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-11-03 22:07:24",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "FLAIROx",
"github_project": "JaxGL",
"github_fetch_exception": true,
"lcname": "jaxgl"
}
