`motrixsim.render`#

Exceptions:

`InvalidArgumentError`	Raise when the given arguments are invalid
`InvalidFileError`	Raise when a file to be loaded is invalid
`OtherRenderError`	Other render errors
`RenderClosedError`	Raise when the render app is closed
`RenderError`	Base class for all render errors

Classes:

`CaptureTask`
`Color`	A color in RGBA format, where each component is a float in the range [0.0, 1.0].
`Image`
`Input`	The Input object represents the input events of the render app.
`Layout`([left, right, top, bottom, width, height])	Layout configuration for camera viewport widgets.
`RenderApp`([log_level, headless, fps])	The RenderApp class is responsible for rendering the simulation scene.
`RenderCamera`
`RenderGizmos`	Gizmos module for rendering simple shapes in immediate mode.
`RenderOpt`	The RenderOpt object represents the options of the render app.
`RenderSettings`(simplify_render_mesh, ...[, ...])	The global render settings for render app.
`RenderUI`	The RenderUI object represents the user interface of the render app.
`RenderWidgets`	Widgets module for creating UI widgets in the render window.
`SystemCamera`

exception motrixsim.render.InvalidArgumentError#

基类：RenderError

Raise when the given arguments are invalid

exception motrixsim.render.InvalidFileError#

基类：RenderError

Raise when a file to be loaded is invalid

exception motrixsim.render.OtherRenderError#

基类：RenderError

Other render errors

exception motrixsim.render.RenderClosedError#

基类：RenderError

Raise when the render app is closed

exception motrixsim.render.RenderError#

基类：Exception

Base class for all render errors

class motrixsim.render.CaptureTask#

基类：object

Methods:

take_image()

Try take the image from the capture task.

Attributes:

state

The state of the capture task.

take_image()#

Try take the image from the capture task.

返回:: The captured image. - If the capture is successful, it will return the image and close the task. - Otherwise, it will return None and the state will remain unchanged.
返回类型:: Option[Image]
抛出:: PyRuntimeError -- If the task is closed or some internal error happend.

state#

The state of the capture task.

pending if the capture is processing.
done if the capture was completed. In this status, you can retrieve the captured result.
closed if the result has been received or the task has been dropped by remote.

Type:: str

class motrixsim.render.Color#

基类：object

A color in RGBA format, where each component is a float in the range [0.0, 1.0].

Methods:

`rgb`(r, g, b)	Create a color with red, green, blue.
`white`()

Attributes:

`a`	The alpha component of the color, representing transparency.
`b`	The blue component of the color.
`g`	The green component of the color.
`r`	The red component of the color.

static rgb(r, g, b)#

Create a color with red, green, blue. Each value should be in the range [0.0, 1.0].

参数:

r (float) -- Red component.
g (float) -- Green component.
b (float) -- Blue component.

返回:

The created color.

返回类型:

Color

static white()#

a#

The alpha component of the color, representing transparency.

Type:: float

b#

The blue component of the color.

Type:: float

g#

The green component of the color.

Type:: float

r#

The red component of the color.

Type:: float

class motrixsim.render.Image#

基类：object

Methods:

save_to_disk(path)

Save the image to disk with the specified path.

Attributes:

pixels

Get the pixels of the image as a numpy array.

save_to_disk(path)#

Save the image to disk with the specified path.

抛出:: PyIOError -- If saving the image fails.

pixels#

Get the pixels of the image as a numpy array.

备注

For regular images, returns a 3D array with shape (height, width, channels) where channels is 3 (RGB) or 4 (RGBA) depending on the underlying image format. For per-instance atlas images, returns a 4D array with shape (num_instances, tile_height, tile_width, channels).

Type:: NDArray[byte]

class motrixsim.render.Input#

基类：object

The Input object represents the input events of the render app.

This class provides access to the input events of the render app. It allows you to check if a key or mouse button is pressed, and get the mouse ray.

Methods:

`is_ctrl_clicked`(ctrl_id)	Check whether a control with the given ID is clicked.
`is_key_just_pressed`(key)	Check if a key is just pressed.
`is_key_pressed`(key)	Check if a key is just pressed.
`is_mouse_just_pressed`(mouse)	Check if a mouse button is just pressed.
`mouse_ray`()	Returns a ray from camera to mouse click position.

is_ctrl_clicked(ctrl_id)#

Check whether a control with the given ID is clicked.

参数:: ctrl_id (int) -- The ID of the control.
返回:: True if the control is clicked, False otherwise.
返回类型:: bool

is_key_just_pressed(key)#

Check if a key is just pressed. Check by inputting the lowercase form of the keyboard keys, such as 'a' 's' 'w' 'd' 'f5' and 'esc'. The details can be found in the "Supported Keyboard Keys Table" under the "IO Input Events" section of the 🎨 渲染器.

参数:: key (str) -- The key to check.
返回:: True if the key is pressed, False otherwise.
返回类型:: bool

is_key_pressed(key)#

Check if a key is just pressed. Check by inputting the lowercase form of the keyboard keys, such as 'a' 's' 'w' 'd' 'f5' and 'esc'. The details can be found in the "Supported Keyboard Keys Table" under the "IO Input Events" section of the 🎨 渲染器.

参数:: key (str) -- The key to check.
返回:: True if the key is pressed, False otherwise.
返回类型:: bool

is_mouse_just_pressed(mouse)#

Check if a mouse button is just pressed.

参数:: mouse (str) -- The mouse button to check.
返回:: True if the mouse button is pressed, False otherwise.
返回类型:: bool

mouse_ray()#

Returns a ray from camera to mouse click position.

返回:: The ray from camera to mouse click position. It is a 6-element array: - Origin : (array[0],array[1],array[2]) - Direction : (array[3],array[4],array[5])
返回类型:: List[float]

class motrixsim.render.Layout(left=None, right=None, top=None, bottom=None, width=None, height=None)#

基类：object

Layout configuration for camera viewport widgets.

Attributes:

`bottom`
`height`
`left`
`right`
`top`
`width`

bottom#

height#

left#

right#

top#

width#

class motrixsim.render.RenderApp(log_level='WARN', headless=False, fps=None)#

基类：object

The RenderApp class is responsible for rendering the simulation scene.

This class provides functionality to load models, update their transformations, and render the scene. It also handles the creation and management of the render application.

Methods:

`create_image`(pixels[, is_srgb, keep_in_cpu])	Create an image from a numpy array of pixel data.
`get_camera`(index)	Get a render camera instance.
`get_texture_image`(name)	Get the image handle of a texture loaded by name.
`launch`(model[, batch, render_offset, ...])	Load a model into the render app.
`set_all_scene_vis`(visible)
`set_main_camera`(camera)	Set the main camera of the render app.
`set_scene_vis`(indices, visible)
`sync`(data[, wait])	Synchronize the render app backend with python data.

Attributes:

`gizmos`	The gizmos module of the render app.
`input`	The input module of the render app.
`is_closed`	Check if the render app is closed.
`opt`	The options of the render app.
`system_camera`	The system camera
`ui`	The UI module of the render app.

create_image(pixels, is_srgb=True, keep_in_cpu=True)#

Create an image from a numpy array of pixel data.

参数:

pixels (NDArray[uint8]) -- A 3D numpy array with shape (height, width, channels) where channels is either 3 (RGB) or 4 (RGBA). The array must be contiguous and contain uint8 values in the range [0, 255].
is_srgb (bool) -- Whether the image is in sRGB color space. Default is True.
keep_in_cpu (bool) -- Whether to keep the image data in CPU memory after uploading to GPU. If False, pixel data cannot be accessed after upload. Default is True.

返回:

A handle to the created image asset that can be used for textures,: materials, etc.

返回类型:

Image

抛出:

InvalidArgumentError -- If the array shape is invalid or not contiguous.
OtherRenderError -- If image creation fails.

get_camera(index)#: Get a render camera instance.

get_texture_image(name)#

Get the image handle of a texture loaded by name.

This is typically used to access TextureSource.Pixels textures at runtime so their pixel data can be read or updated.

参数:

name (str) -- The texture name as defined in the MSD assets.

返回:

A handle to the texture's image. Updating its pixels will update the: rendered texture in the scene.

返回类型:

Image

抛出:

InvalidArgumentError -- If no texture with the given name was found.

launch(model, batch=1, render_offset=None, render_settings=Ellipsis)#

Load a model into the render app.

参数:

model (SceneModel) -- The scene model to load into the render app.
batch (int, optional) -- The number of instances to create. Default is 1.
render_offset (Optional[List[List[float]]], optional) -- The offset of each instance in render space. Default is None.

抛出:

RenderClosedError -- If the render app is closed.
InvalidArgumentError -- If the file is invalid.
InvalidFileError -- If there are issues with file operations.
OtherRenderError -- For other unexpected errors.

set_all_scene_vis(visible)#

set_main_camera(camera)#

Set the main camera of the render app.

参数:

camera (Optional[Camera]) -- The camera to set as the main camera. If None, the system
used. (camera will be)

set_scene_vis(indices, visible)#

sync(data, wait=False)#

Synchronize the render app backend with python data.

This is the explicit synchronization point between the Python client loop and the render service. Call it at render cadence, after one or more physics steps, rather than from every physics step in a hot loop. Calling it more frequently than the service can consume frames will throttle the caller.

Each successful call also drains service-to-client events queued since the previous sync call. Window input, UI events, and camera updates are only delivered to the Python side when sync is called often enough to consume them.

参数:

data (None | SceneData | NDArray) --
The scene data to synchronize with the renderer.

This argument accept following types:
- None: No data is provided, the render will use the last known transforms.
- SceneData: The scene data used to update the render. If you lanched the render with repeat > 1, the shape the of data must be (repeat,).
- NDArray: A 3D numpy array with shape (num_instances, num_links, 7), where the last dimension represents the pose of each link in the format [x, y, z, qx, qy, qz, qw].
wait (bool) -- If True, block until all pending capture tasks are completed. Default is False.

抛出:

RenderClosedError -- If the render app is closed.
InvalidArgumentError -- If neither datas nor poses is provided correctly.
InvalidFileError -- If there are issues with file operations.
OtherRenderError -- Not launched with model file.

gizmos#

The gizmos module of the render app.

Return the gizmos module for rendering simple shapes in immediate mode.

Type:: RenderGizmos

input#

The input module of the render app.

Return the input module for handling user input events.

Type:: Input

is_closed#: Check if the render app is closed.

opt#

The options of the render app.

Return the options of the render app, which can be used to configure various settings.

Type:: RenderOpt

system_camera#: The system camera

ui#

The UI module of the render app.

Type:: RenderUI

class motrixsim.render.RenderCamera#

基类：object

Methods:

capture()

Request a capture from this camera.

Attributes:

active

Activates or deactivates the camera.

capture()#

Request a capture from this camera. This operation is asynchronous and you need to query the result by the returned CaptureTask.

返回:: An async task that can be used to check the capture state and get the captured image.
返回类型:: CaptureTask

active#: Activates or deactivates the camera.

class motrixsim.render.RenderGizmos#

基类：object

Gizmos module for rendering simple shapes in immediate mode.

Methods:

`draw_arrow`(start, end[, color])	Draw a arrow.
`draw_axes`([pos, rot, length])	Draw the XYZ axes at the given position and rotation.
`draw_capsule`(half_height, radius, pos, rot)	Draw a capsule.
`draw_cuboid`(size, pos, rot[, color])	Draw a cuboid.
`draw_cylinder`(half_height, radius, pos, rot)	Draw a cylinder.
`draw_grid`(pos, rot[, x_count, y_count, ...])	Draw a rectangle.
`draw_line`(start, end[, color])	Draw a line.
`draw_ray`(start, vector[, color])	Draw a ray.
`draw_rect`(width, height, pos, rot[, color])	Draw a rectangle.
`draw_sphere`(radius, pos[, color])	Draw a sphere at the given position with the specified radius and color.

Attributes:

`collider_color`	Set the color of collider gizmos.
`draw_collider`	Enable or disable collider gizmos.
`draw_convex_collider`	Enable or disable drawing mesh collider gizmos from runtime convex hulls.
`draw_joint`	Enable or disable joint gizmos.
`draw_site`	Enable or disable site gizmos.
`joint_color`	Set the color of joint gizmos.
`joint_size`	Set the size of joint gizmos.
`line_width`	Set the line width of all gizmos.

draw_arrow(start, end, color=Ellipsis)#

Draw a arrow.

参数:

start (list[float] | NDArray) -- The start point of the arrow in 3D space.
end (list[float] | NDArray) -- The end point of the arrow.
color (Color) -- Color of the arrow.

draw_axes(pos=None, rot=None, length=1.0)#

Draw the XYZ axes at the given position and rotation.

参数:

pos (list[float] | ndarray, optional) -- the position of axes with (x,y,z) format.
rot (list[float] | ndarray, optional) -- the rotation of axes as a quaternion (x, y, z,
length (float) -- the length of each axis.

w).

draw_capsule(half_height, radius, pos, rot, color=Ellipsis)#

Draw a capsule.

参数:

half_height (float) -- The half height of the capsule.
radius (float) -- The radius of the capsule.
pos (list[float] | NDArray) -- Position of the capsule in 3D space.
rot (list[float] | NDArray) -- Rotation of the capsule as a quaternion (x, y, z, w).
color (Color) -- Color of the capsule.

draw_cuboid(size, pos, rot, color=Ellipsis)#

Draw a cuboid.

参数:

size (list[float] | NDArray) -- Size of the cuboid in 3D space.
pos (list[float] | NDArray) -- Position of the cuboid in 3D space.
rot (list[float] | NDArray) -- Rotation of the cuboid as a quaternion (x, y, z, w).
color (Color) -- Color of the cuboid.

draw_cylinder(half_height, radius, pos, rot, color=Ellipsis)#

Draw a cylinder.

参数:

half_height (float) -- The half height of the cylinder.
radius (float) -- The radius of the cylinder.
pos (list[float] | NDArray) -- Position of the cylinder in 3D space.
rot (list[float] | NDArray) -- Rotation of the cylinder as a quaternion (x, y, z, w).
color (Color) -- Color of the cylinder.

draw_grid(pos, rot, x_count=2, y_count=2, z_count=2, spacing=None, color=Ellipsis)#

Draw a rectangle.

参数:

pos (list[float] | NDArray) -- Position of the rectangle in 3D space.
rot (list[float] | NDArray) -- Rotation of the rectangle as a quaternion (x, y, z, w).
x_count (unsigned int) -- The number of grid cells along the x-axis.
y_count (unsigned int) -- The number of grid cells along the y-axis.
z_count (unsigned int) -- The number of grid cells along the z-axis.
spacing (list[float] | NDArray) -- The spacing(x,y,z) between the grid cells.
color (Color) -- Color of the rectangle.

draw_line(start, end, color=Ellipsis)#

Draw a line.

参数:

start (list[float] | NDArray) -- The start point of the line in 3D space.
end (list[float] | NDArray) -- The end point of the line.
color (Color) -- Color of the line.

draw_ray(start, vector, color=Ellipsis)#

Draw a ray.

参数:

start (list[float] | NDArray) -- The start point of the ray in 3D space.
vector (list[float] | NDArray) -- The direction of the ray.
color (Color) -- Color of the ray.

draw_rect(width, height, pos, rot, color=Ellipsis)#

Draw a rectangle.

参数:

width (float) -- The width of the rectangle.
height (float) -- The height of the rectangle.
pos (list[float] | NDArray) -- Position of the rectangle in 3D space.
rot (list[float] | NDArray) -- Rotation of the rectangle as a quaternion (x, y, z, w).
color (Color) -- Color of the rectangle.

draw_sphere(radius, pos, color=Ellipsis)#

Draw a sphere at the given position with the specified radius and color.

参数:

radius (float) -- The radius of the sphere.
pos (list[float] | NDArray) -- The position of the sphere in 3D space.
color (Color) -- The color of the sphere.

collider_color#

Set the color of collider gizmos.

参数:: color (Color) -- Color of the collider gizmos.

draw_collider#

Enable or disable collider gizmos.

参数:: flag (bool) -- Enable/Disable the collider gizmos.

draw_convex_collider#

Enable or disable drawing mesh collider gizmos from runtime convex hulls.

参数:: flag (bool) -- Enable/Disable runtime convex hull collider gizmos.

draw_joint#

Enable or disable joint gizmos.

参数:: flag (bool) -- Enable/Disable the joint gizmos.

draw_site#

Enable or disable site gizmos.

参数:: flag (bool) -- Enable/Disable the site gizmos.

joint_color#

Set the color of joint gizmos.

参数:: color (Color) -- Color of the joint gizmos.

joint_size#

Set the size of joint gizmos.

参数:: size (float) -- The size of joint gizmos.

line_width#

Set the line width of all gizmos.

参数:: width (float) -- The line width of all gizmos.

class motrixsim.render.RenderOpt#

基类：object

The RenderOpt object represents the options of the render app.

Methods:

`is_left_panel_vis`()	Check if the left panel is visible.
`set_group_vis`(group, visible)	Set the visibility of a geometry group.
`set_left_panel_vis`(enabled)	Set the visibility of the left panel.

is_left_panel_vis()#

Check if the left panel is visible.

返回:: True if the left panel is visible, False otherwise.
返回类型:: bool

set_group_vis(group, visible)#

Set the visibility of a geometry group.

参数:

group (int) -- The geometry group index (0-MAX_GROUP_SIZE-1).
visible (bool) -- True to make the group visible, False to hide it.

抛出:

Exception -- If group index is out of valid range.

set_left_panel_vis(enabled)#

Set the visibility of the left panel.

参数:: enabled (bool) -- True to show the left panel, False to hide it.

class motrixsim.render.RenderSettings(simplify_render_mesh, enable_shadow, enable_ssao, enable_oit, share_lights_between_envs, enable_ssgi=False)#

基类：object

The global render settings for render app.

Methods:

`performance`()	Get a render settings with high performance.
`quality`()	Get a render settings with high quality.

Attributes:

`enable_oit`	Whether to enable OIT(Order Independent Transparency) effect.
`enable_shadow`	Whether to render shadows for lights.
`enable_ssao`	Whether to enable SSAO(Screen Space Ambient Occlusion) effect.
`enable_ssgi`	Whether to enable SSGI(Screen Space Global Illumination) effect.
`share_lights_between_envs`	Whether to share lights between multiple simulation worlds, for performance optimization.
`simplify_render_mesh`	Wether to simplify render mesh when loading the model.

static performance()#: Get a render settings with high performance.

static quality()#: Get a render settings with high quality.

enable_oit#: Whether to enable OIT(Order Independent Transparency) effect.

enable_shadow#: Whether to render shadows for lights.

enable_ssao#: Whether to enable SSAO(Screen Space Ambient Occlusion) effect.

enable_ssgi#: Whether to enable SSGI(Screen Space Global Illumination) effect.

share_lights_between_envs#: Whether to share lights between multiple simulation worlds, for performance optimization. If false, each simulation world will have its own set of lights, which may cost more rendering resources, or out of memory crash depends on the number of simulation worlds and scene complexity.

simplify_render_mesh#: Wether to simplify render mesh when loading the model.

class motrixsim.render.RenderUI#

基类：object

The RenderUI object represents the user interface of the render app.

This class provides access to the user interface of the render app. It allows you to add buttons, toggles, and other UI elements to the render app.

Methods:

`add_button`(label, on_click)	Add a button to the user interface.
`add_toggle`(label, default, on_changed)	Add a toggle to the user interface.

add_button(label, on_click)#

Add a button to the user interface.

参数:

label (str) -- The label of the button.
on_click (PyAny) -- The callback function to be called when the button is clicked.

返回:

The ID of the button.

返回类型:

int

add_toggle(label, default, on_changed)#

Add a toggle to the user interface.

参数:

label (str) -- The label of the toggle.
default (bool) -- The default state of the toggle.
on_changed (PyAny) -- The callback function to be called when the toggle is changed.

class motrixsim.render.RenderWidgets#

基类：object

Widgets module for creating UI widgets in the render window.

Methods:

`create_camera_viewport`(camera[, layout, ...])	Create a camera viewport widget in the render window.
`create_image_widget`(image[, layout])	Create an image widget in the render window.

create_camera_viewport(camera, layout=None, sim_world_index=0)#

Create a camera viewport widget in the render window.

备注

This creates a camera viewport overlay that displays the output of a sensor camera. The layout parameters accept multiple formats:

String: "50px" for pixels, "50%" for percentage, "auto" for automatic

Number: Interpreted as pixels (e.g., 50 = 50px)

参数:

camera (Camera) -- The camera object to display in the viewport.
layout (Optional[Layout]) -- The layout configuration for the viewport. If None, uses default layout (50px from top-left with 200x200 size).
sim_world_index (int) -- The index of the simulation world. Default is 0.

Example:

renderer = motrixsim.render.RenderApp()
renderer.launch(model)
cameras = model.cameras
# Create viewport with default layout
vp = renderer.widgets.create_camera_viewport(cameras[0])
# Create viewport with custom layout
layout = motrixsim.render.Layout(left=50, top=50, width=400, height=300)
vp = renderer.widgets.create_camera_viewport(cameras[0], layout=layout)
# Create viewport with percentage-based layout
layout = motrixsim.render.Layout(
    left="10%", top="10%", width="50%", height="50%"
)
vp = renderer.widgets.create_camera_viewport(cameras[1], sim_world_index=1,
    layout=layout)
# Access camera properties
print(f"Camera name: {vp.camera.name}")

create_image_widget(image, layout=None)#

Create an image widget in the render window.

备注

This creates an image widget overlay that displays the specified image. The layout parameters accept multiple formats: - String: "50px" for pixels, "50%" for percentage, "auto" for automatic - Number: Interpreted as pixels (e.g., 50 = 50px)

参数:

image (Image) -- The image object to display in the widget.
layout (Optional[Layout]) -- The layout configuration for the widget. If None, uses default layout (0px from top-left with automatic size).

Example:

renderer = motrixsim.render.RenderApp()
renderer.launch(model)
# Load an image
img = motrixsim.render.Image.from_file("path/to/image.png")
# Create widget with default layout
widget = renderer.widgets.create_image_widget(img)
# Create widget with custom layout
layout = motrixsim.render.Layout(left=100, top=100, width=200, height=200)
widget = renderer.widgets.create_image_widget(img, layout=layout)
# Update the widget
widget.update(layout=new_layout)

class motrixsim.render.SystemCamera#

基类：object

Methods:

`capture`()	Request a capture from the system camera.
`set_pose`([position, rotation])	Set the global pose of the system camera.
`set_view`(lookat, distance, elevation, azimuth)	Set the system camera from orbit/look-at view parameters.

Attributes:

`active`
`pose`	The global pose of the system camera as a 7 element array with the format [pos_x, pos_y, pos_z, rot_i, rot_j, rot_k, rot_w]

capture()#

Request a capture from the system camera.

返回:: An async task that can be used to check the capture state and get the captured image.
返回类型:: CaptureTask

set_pose(position=None, rotation=None)#

Set the global pose of the system camera.

参数:

position (Optional[Sequence[float]]) -- Optional world position [x, y, z].
rotation (Optional[Sequence[float]]) -- Optional world rotation quaternion [i, j, k, w].

set_view(lookat, distance, elevation, azimuth)#

Set the system camera from orbit/look-at view parameters.

The camera looks at lookat from the given orbit distance and angles. The renderer applies the corresponding system camera transform and synchronizes the interactive orbit camera state, so subsequent mouse drags continue from the same view. This camera is the renderer system camera, not a model-defined MJCF camera.

参数:

lookat (Sequence[float]) -- World-space target point [x, y, z].
distance (float) -- Distance from the camera to lookat.
elevation (float) -- Elevation angle in degrees. Negative values place the camera above lookat and look down; positive values place it below lookat and look up.
azimuth (float) -- Azimuth angle around the world Z axis in degrees.

active#

pose#

The global pose of the system camera as a 7 element array with the format [pos_x, pos_y, pos_z, rot_i, rot_j, rot_k, rot_w]

Type:: NDArray[float]

motrixsim.render#

`motrixsim.render`#