API 参考：场景（scene）¶

共 10 个类，按字母排序。每个类含中文说明、继承链、参数表与上手示例，API 文档由 autodoc 从 manim v0.21.0 源码自动生成；带完整中文精讲的类见 API 索引。

DefaultSectionType¶

章节类型枚举：normal/skip/speedup，控制分节导出时该节的处理方式。

API 文档¶

class manim.DefaultSectionType(*values)¶

基类：StrEnum

The type of a section can be used for third party applications. A presentation system could for example use the types to created loops.

Examples¶

This class can be reimplemented for more types:

class PresentationSectionType(str, Enum):
    # start, end, wait for continuation by user
    NORMAL = "presentation.normal"
    # start, end, immediately continue to next section
    SKIP = "presentation.skip"
    # start, end, restart, immediately continue to next section when continued by user
    LOOP = "presentation.loop"
    # start, end, restart, finish animation first when user continues
    COMPLETE_LOOP = "presentation.complete_loop"

LinearTransformationScene¶

线性变换演示场景：自带基向量网格与 apply_matrix 快捷方法，线代可视化开箱即用。

继承关系¶

参数¶

`include_background_plane`	True，bool
`include_foreground_plane`	True，bool
`background_plane_kwargs`	None，dict[str, Any] \| None
`foreground_plane_kwargs`	None，dict[str, Any] \| None
`show_coordinates`	False，bool
`show_basis_vectors`	True，bool
`basis_vector_stroke_width`	6，float
`i_hat_color`	ManimColor('#83C167')，ParsableManimColor
`j_hat_color`	ManimColor('#FC6255')，ParsableManimColor
`leave_ghost_vectors`	False，bool

快速上手¶

class Demo(LinearTransformationScene):
    def construct(self):
        self.apply_matrix([[1, 1], [0, 1]])

API 文档¶

基类：VectorScene

This scene contains special methods that make it especially suitable for showing linear transformations.

Parameters¶

include_background_plane: Whether or not to include the background plane in the scene.
include_foreground_plane: Whether or not to include the foreground plane in the scene.
background_plane_kwargs: Parameters to be passed to NumberPlane to adjust the background plane.
foreground_plane_kwargs: Parameters to be passed to NumberPlane to adjust the foreground plane.
show_coordinates: Whether or not to include the coordinates for the background plane.
show_basis_vectors: Whether to show the basis x_axis -> i_hat and y_axis -> j_hat vectors.
basis_vector_stroke_width: The stroke_width of the basis vectors.
i_hat_color: The color of the i_hat vector.
j_hat_color: The color of the j_hat vector.
leave_ghost_vectors: Indicates the previous position of the basis vectors following a transformation.

Examples¶

class LinearTransformationSceneExample(LinearTransformationScene):
    def __init__(self, **kwargs):
        LinearTransformationScene.__init__(
            self,
            show_coordinates=True,
            leave_ghost_vectors=True,
            **kwargs
        )

    def construct(self):
        matrix = [[1, 1], [0, 1]]
        self.apply_matrix(matrix)
        self.wait()

add_background_mobject(*mobjects: Mobject) → None¶

Adds the mobjects to the special list self.background_mobjects.

Parameters¶

*mobjects: The mobjects to add to the list.

add_foreground_mobject(*mobjects: Mobject) → None¶

Adds the mobjects to the special list self.foreground_mobjects.

Parameters¶

*mobjects: The mobjects to add to the list

add_moving_mobject(mobject: Mobject, target_mobject: Mobject | None = None) → None¶

Adds the mobject to the special list self.moving_mobject, and adds a property to the mobject called mobject.target, which keeps track of what the mobject will move to or become etc.

Parameters¶

mobject: The mobjects to add to the list
target_mobject: What the moving_mobject goes to, etc.

add_special_mobjects(mob_list: list[Mobject], *mobs_to_add: Mobject) → None¶

Adds mobjects to a separate list that can be tracked, if these mobjects have some extra importance.

Parameters¶

mob_list: The special list to which you want to add these mobjects.
*mobs_to_add: The mobjects to add.

add_title(title: str | MathTex | Tex, scale_factor: float = 1.5, animate: bool = False) → Self¶

Adds a title, after scaling it, adding a background rectangle, moving it to the top and adding it to foreground_mobjects adding it as a local variable of self. Returns the Scene.

Parameters¶

title: What the title should be.
scale_factor: How much the title should be scaled by.
animate: Whether or not to animate the addition.

Returns¶

LinearTransformationScene: The scene with the title added to it.

add_transformable_label(vector: Vector, label: MathTex | str, transformation_name: str | MathTex = 'L', new_label: str | MathTex | None = None, **kwargs: Any) → MathTex¶

Method for creating, and animating the addition of a transformable label for the vector.

Parameters¶

vector: The vector for which the label must be added.
label: The MathTex/string of the label.
transformation_name: The name to give the transformation as a label.
new_label: What the label should display after a Linear Transformation
**kwargs: Any valid keyword argument of get_vector_label

Returns¶

MathTex: The MathTex of the label.

add_transformable_mobject(*mobjects: Mobject) → None¶

Adds the mobjects to the special list self.transformable_mobjects.

Parameters¶

*mobjects: The mobjects to add to the list.

add_unit_square(animate: bool = False, **kwargs: Any) → Self¶

Adds a unit square to the scene via self.get_unit_square.

Parameters¶

animate: Whether or not to animate the addition with DrawBorderThenFill.
**kwargs: Any valid keyword arguments of self.get_unit_square()

Returns¶

Square: The unit square.

Adds a vector to the scene, and puts it in the special list self.moving_vectors.

Parameters¶

vector: It can be a pre-made graphical vector, or the coordinates of one.
color: The string of the hex color of the vector. This is only taken into consideration if 'vector' is not an Arrow. Defaults to YELLOW.
**kwargs: Any valid keyword argument of VectorScene.add_vector.

Returns¶

Arrow: The arrow representing the vector.

apply_function(function: MappingFunction, added_anims: list[Animation] = [], **kwargs: Any) → None¶

Applies the given function to each of the mobjects in self.transformable_mobjects, and plays the animation showing this.

Parameters¶

function: The function that affects each point of each mobject in self.transformable_mobjects.
added_anims: Any other animations that need to be played simultaneously with this.
**kwargs: Any valid keyword argument of a self.play() call.

apply_inverse(matrix: ndarray | list | tuple, **kwargs: Any) → None¶

This method applies the linear transformation represented by the inverse of the passed matrix to the number plane, and each vector/similar mobject on it.

Parameters¶

matrix: The matrix whose inverse is to be applied.
**kwargs: Any valid keyword argument of self.apply_matrix()

apply_inverse_transpose(t_matrix: ndarray | list | tuple, **kwargs: Any) → None¶

Applies the inverse of the transformation represented by the given transposed matrix to the number plane and each vector/similar mobject on it.

Parameters¶

t_matrix: The matrix.
**kwargs: Any valid keyword argument of self.apply_transposed_matrix()

apply_matrix(matrix: ndarray | list | tuple, **kwargs: Any) → None¶

Applies the transformation represented by the given matrix to the number plane, and each vector/similar mobject on it.

Parameters¶

matrix: The matrix.
**kwargs: Any valid keyword argument of self.apply_transposed_matrix()

apply_nonlinear_transformation(function: Callable[[ndarray], ndarray], **kwargs: Any) → None¶

Applies the non-linear transformation represented by the given function to the number plane and each vector/similar mobject on it.

Parameters¶

function: The function.
**kwargs: Any valid keyword argument of self.apply_function()

apply_transposed_matrix(transposed_matrix: ndarray | list | tuple, **kwargs: Any) → None¶

Applies the transformation represented by the given transposed matrix to the number plane, and each vector/similar mobject on it.

Parameters¶

transposed_matrix: The matrix.
**kwargs: Any valid keyword argument of self.apply_function()

get_ghost_vectors() → VGroup¶: Returns all ghost vectors ever added to self. Each element is a VGroup of two ghost vectors.

get_matrix_transformation(matrix: np.ndarray | list | tuple) → Callable[[Point3D], Point3D]¶

Returns a function corresponding to the linear transformation represented by the matrix passed.

Parameters¶

matrix: The matrix.

get_moving_mobject_movement(func: MappingFunction) → Transform¶

This method returns an animation that moves a mobject in "self.moving_mobjects" to its corresponding .target value. func is a function that determines where the .target goes.

Parameters¶

func: The function that determines where the .target of the moving mobject goes.

Returns¶

Animation: The animation of the movement.

get_piece_movement(pieces: Iterable[Mobject]) → Transform¶

This method returns an animation that moves an arbitrary mobject in "pieces" to its corresponding .target value. If self.leave_ghost_vectors is True, ghosts of the original positions/mobjects are left on screen

Parameters¶

pieces: The pieces for which the movement must be shown.

Returns¶

Animation: The animation of the movement.

get_transformable_label_movement() → Transform¶

This method returns an animation that moves all labels in "self.transformable_labels" to its corresponding .target .

Returns¶

Animation: The animation of the movement.

get_transposed_matrix_transformation(transposed_matrix: np.ndarray | list | tuple) → Callable[[Point3D], Point3D]¶

Returns a function corresponding to the linear transformation represented by the transposed matrix passed.

Parameters¶

transposed_matrix: The matrix.

Returns a unit square for the current NumberPlane.

Parameters¶

color: The string of the hex color code of the color wanted.
opacity: The opacity of the square
stroke_width: The stroke_width in pixels of the border of the square

Returns¶

Square

get_vector_movement(func: MappingFunction) → Transform¶

This method returns an animation that moves a mobject in "self.moving_vectors" to its corresponding .target value. func is a function that determines where the .target goes.

Parameters¶

func: The function that determines where the .target of the moving mobject goes.

Returns¶

Animation: The animation of the movement.

setup() → None¶: This is meant to be implemented by any scenes which are commonly subclassed, and have some common setup involved before the construct method is called.

write_vector_coordinates(vector: Vector, **kwargs: Any) → Matrix¶

Returns a column matrix indicating the vector coordinates, after writing them to the screen, and adding them to the special list self.foreground_mobjects

Parameters¶

vector: The arrow representing the vector.
**kwargs: Any valid keyword arguments of VectorScene.write_vector_coordinates

Returns¶

Matrix: The column matrix representing the vector.

MovingCameraScene¶

镜头运动场景：self.camera.frame 自由推拉摇移，适合走遍整张图的讲解。

继承关系¶

参数¶

快速上手¶

self.play(self.camera.frame.animate.scale(0.5).shift(RIGHT))

API 文档¶

class manim.MovingCameraScene(camera_class: type[Camera] = <class 'manim.camera.moving_camera.MovingCamera'>, **kwargs: Any)¶

基类：Scene

This is a Scene, with special configurations and properties that make it suitable for cases where the camera must be moved around.

Note: Examples are included in the moving_camera_scene module documentation, see below in the 'see also' section.

参见

moving_camera_scene MovingCamera

get_moving_mobjects(*animations: Animation) → list[Mobject]¶

This method returns a list of all of the Mobjects in the Scene that are moving, that are also in the animations passed.

Parameters¶

*animations: The Animations whose mobjects will be checked.

Scene¶

场景基类：一切动画的主入口。construct() 里用 add/play 编排，掌握 Scene 就掌握了 Manim 的导演椅。

参数¶

`renderer`	None
`camera_class`	`<class 'manim.camera.camera…，type[Camera]`
`always_update_mobjects`	False，bool
`random_seed`	None，int \| None
`skip_animations`	False，bool

快速上手¶

class Demo(Scene):
    def construct(self):
        self.play(Write(Text("Hello")))

API 文档¶

class manim.Scene(renderer: CairoRenderer | OpenGLRenderer | None = None, camera_class: type[Camera] = <class 'manim.camera.camera.Camera'>, always_update_mobjects: bool = False, random_seed: int | None = None, skip_animations: bool = False)¶

基类：object

A Scene is the canvas of your animation.

The primary role of Scene is to provide the user with tools to manage mobjects and animations. Generally speaking, a manim script consists of a class that derives from Scene whose Scene.construct() method is overridden by the user's code.

Mobjects are displayed on screen by calling Scene.add() and removed from screen by calling Scene.remove(). All mobjects currently on screen are kept in Scene.mobjects. Animations are played by calling Scene.play().

A Scene is rendered internally by calling Scene.render(). This in turn calls Scene.setup(), Scene.construct(), and Scene.tear_down(), in that order.

It is not recommended to override the __init__ method in user Scenes. For code that should be ran before a Scene is rendered, use Scene.setup() instead.

Examples¶

Override the Scene.construct() method with your code.

class MyScene(Scene):
    def construct(self):
        self.play(Write(Text("Hello World!")))

add(*mobjects: Mobject | OpenGLMobject) → Self¶

Mobjects will be displayed, from background to foreground in the order with which they are added.

Parameters¶

*mobjects: Mobjects to add.

Returns¶

Scene: The same scene after adding the Mobjects in.

add_foreground_mobject(mobject: Mobject) → Scene¶

Adds a single mobject to the foreground, and internally to the list foreground_mobjects, and mobjects.

Parameters¶

mobject: The Mobject to add to the foreground.

Returns¶

Scene: The Scene, with the foreground mobject added.

add_foreground_mobjects(*mobjects: Mobject) → Scene¶

Adds mobjects to the foreground, and internally to the list foreground_mobjects, and mobjects.

Parameters¶

*mobjects: The Mobjects to add to the foreground.

Returns¶

Scene: The Scene, with the foreground mobjects added.

add_sound(sound_file: str, time_offset: float = 0, gain: float | None = None, **kwargs: Any) → None¶

This method is used to add a sound to the animation.

Parameters¶

sound_file: The path to the sound file.
time_offset: The offset in the sound file after which the sound can be played.
gain: Amplification of the sound.

Examples¶

class SoundExample(Scene):
    # Source of sound under Creative Commons 0 License. https://freesound.org/people/Druminfected/sounds/250551/
    def construct(self):
        dot = Dot().set_color(GREEN)
        self.add_sound("click.wav")
        self.add(dot)
        self.wait()
        self.add_sound("click.wav")
        dot.set_color(BLUE)
        self.wait()
        self.add_sound("click.wav")
        dot.set_color(RED)
        self.wait()

Download the resource for the previous example here .

add_subcaption(content: str, duration: float = 1, offset: float = 0) → None¶

Adds an entry in the corresponding subcaption file at the current time stamp.

The current time stamp is obtained from Scene.time.

Parameters¶

content: The subcaption content.
duration: The duration (in seconds) for which the subcaption is shown.
offset: This offset (in seconds) is added to the starting time stamp of the subcaption.

Examples¶

This example illustrates both possibilities for adding subcaptions to Manimations:

class SubcaptionExample(Scene):
    def construct(self):
        square = Square()
        circle = Circle()

        # first option: via the add_subcaption method
        self.add_subcaption("Hello square!", duration=1)
        self.play(Create(square))

        # second option: within the call to Scene.play
        self.play(
            Transform(square, circle), subcaption="The square transforms."
        )

add_updater(func: Callable[[float], None]) → None¶

Add an update function to the scene.

The scene updater functions are run every frame, and they are the last type of updaters to run.

警告

When using the Cairo renderer, scene updaters that modify mobjects are not detected in the same way that mobject updaters are. To be more concrete, a mobject only modified via a scene updater will not necessarily be added to the list of moving mobjects and thus might not be updated every frame.

TL;DR: Use mobject updaters to update mobjects.

Parameters¶

func: The updater function. It takes a float, which is the time difference since the last update (usually equal to the frame rate).

Parameters¶

*mobjects: The mobject(s) to push to the back of the scene.

Returns¶

Scene: The Scene, with the mobjects pushed to the back of the scene.

bring_to_front(*mobjects: Mobject) → Scene¶

Adds the passed mobjects to the scene again, pushing them to he front of the scene.

Parameters¶

*mobjects: The mobject(s) to bring to the front of the scene.

Returns¶

Scene: The Scene, with the mobjects brought to the front of the scene.

clear() → Self¶

Removes all mobjects present in self.mobjects and self.foreground_mobjects from the scene.

Returns¶

Scene: The Scene, with all of its mobjects in self.mobjects and self.foreground_mobjects removed.

compile_animation_data(*animations: Animation | Mobject | _AnimationBuilder, **play_kwargs: Any) → Self | None¶

Given a list of animations, compile the corresponding static and moving mobjects, and gather the animation durations.

This also begins the animations.

Parameters¶

animations: Animation or mobject with mobject method and params
play_kwargs: Named parameters affecting what was passed in animations, e.g. run_time, lag_ratio and so on.

Returns¶

self, None: None if there is nothing to play, or self otherwise.

compile_animations(*args: Animation | Mobject | _AnimationBuilder, **kwargs: Any) → list[Animation]¶

Creates _MethodAnimations from any _AnimationBuilders and updates animation kwargs with kwargs passed to play().

Parameters¶

*args: Animations to be played.
**kwargs: Configuration for the call to play().

Returns¶

Tuple[Animation]: Animations to be played.

construct() → None¶

Add content to the Scene.

From within Scene.construct(), display mobjects on screen by calling Scene.add() and remove them from screen by calling Scene.remove(). All mobjects currently on screen are kept in Scene.mobjects. Play animations by calling Scene.play().

Notes¶

Initialization code should go in Scene.setup(). Termination code should go in Scene.tear_down().

Examples¶

A typical manim script includes a class derived from Scene with an overridden Scene.construct() method:

class MyScene(Scene):
    def construct(self):
        self.play(Write(Text("Hello World!")))

Parameters¶

*keys: Name(s) of the argument(s) to return the attribute of.

Returns¶

list: List of attributes of the passed identifiers.

get_mobject_family_members() → list[Mobject]¶

Returns list of family-members of all mobjects in scene. If a Circle() and a VGroup(Rectangle(),Triangle()) were added, it returns not only the Circle(), Rectangle() and Triangle(), but also the VGroup() object.

Returns¶

list: List of mobject family members.

get_moving_mobjects(*animations: Animation) → list[Mobject]¶

Gets all moving mobjects in the passed animation(s).

Parameters¶

*animations: The animations to check for moving mobjects.

Returns¶

list: The list of mobjects that could be moving in the Animation(s)

get_restructured_mobject_list(mobjects: Iterable[Mobject], to_remove: Iterable[Mobject]) → list[Mobject]¶

Given a list of mobjects and a list of mobjects to be removed, this filters out the removable mobjects from the list of mobjects.

Parameters¶

mobjects: The Mobjects to check.
to_remove: The list of mobjects to remove.

Returns¶

list: The list of mobjects with the mobjects to remove removed.

get_run_time(animations: list[Animation]) → float¶

Gets the total run time for a list of animations.

Parameters¶

animations: A list of the animations whose total run_time is to be calculated.

Returns¶

float: The total run_time of all of the animations in the list.

get_time_progression(run_time: float, description: str, n_iterations: int | None = None, override_skip_animations: bool = False) → tqdm[float]¶

You will hardly use this when making your own animations. This method is for Manim's internal use.

Returns a CommandLine ProgressBar whose fill_time is dependent on the run_time of an animation, the iterations to perform in that animation and a bool saying whether or not to consider the skipped animations.

Parameters¶

run_time: The run_time of the animation.
n_iterations: The number of iterations in the animation.
override_skip_animations: Whether or not to show skipped animations in the progress bar.

Returns¶

time_progression: The CommandLine Progress Bar.

get_top_level_mobjects() → list[Mobject]¶

Returns all mobjects which are not submobjects.

Returns¶

list: List of top level mobjects.

interactive_embed() → None¶: Like embed(), but allows for screen interaction.

is_current_animation_frozen_frame() → bool¶: Returns whether the current animation produces a static frame (generally a Wait).

next_section(name: str = 'unnamed', section_type: str = DefaultSectionType.NORMAL, skip_animations: bool = False) → None¶: Create separation here; the last section gets finished and a new one gets created. skip_animations skips the rendering of all animations in this section. Refer to the documentation</tutorials/output_and_config> on how to use sections.

pause(duration: float = 1.0) → None¶

Pauses the scene (i.e., displays a frozen frame).

This is an alias for wait() with frozen_frame set to True.

Parameters¶

duration: The duration of the pause.

Parameters¶

args: Animations to be played.
subcaption: The content of the external subcaption that should be added during the animation.
subcaption_duration: The duration for which the specified subcaption is added. If None (the default), the run time of the animation is taken.
subcaption_offset: An offset (in seconds) for the start time of the added subcaption.
kwargs: All other keywords are passed to the renderer.

play_internal(skip_rendering: bool = False) → None¶

This method is used to prep the animations for rendering, apply the arguments and parameters required to them, render them, and write them to the video file.

Parameters¶

skip_rendering: Whether the rendering should be skipped, by default False

remove(*mobjects: Mobject) → Self¶

Removes mobjects in the passed list of mobjects from the scene and the foreground, by removing them from "mobjects" and "foreground_mobjects"

Parameters¶

*mobjects: The mobjects to remove.

remove_foreground_mobject(mobject: Mobject) → Scene¶

Removes a single mobject from the foreground, and internally from the list foreground_mobjects.

Parameters¶

mobject: The mobject to remove from the foreground.

Returns¶

Scene: The Scene, with the foreground mobject removed.

remove_foreground_mobjects(*to_remove: Mobject) → Scene¶

Removes mobjects from the foreground, and internally from the list foreground_mobjects.

Parameters¶

*to_remove: The mobject(s) to remove from the foreground.

Returns¶

Scene: The Scene, with the foreground mobjects removed.

remove_updater(func: Callable[[float], None]) → None¶

Remove an update function from the scene.

Parameters¶

func: The updater function to be removed.

Parameters¶

preview: If true, opens scene in a file viewer.

replace(old_mobject: Mobject, new_mobject: Mobject) → None¶

Replace one mobject in the scene with another, preserving draw order.

If old_mobject is a submobject of some other Mobject (e.g. a Group), the new_mobject will replace it inside the group, without otherwise changing the parent mobject.

Parameters¶

old_mobject: The mobject to be replaced. Must be present in the scene.
new_mobject: A mobject which must not already be in the scene.

restructure_mobjects(to_remove: Sequence[Mobject], mobject_list_name: str = 'mobjects', extract_families: bool = True) → Scene¶

tl:wr: If your scene has a Group(), and you removed a mobject from the Group, this dissolves the group and puts the rest of the mobjects directly in self.mobjects or self.foreground_mobjects.

In cases where the scene contains a group, e.g. Group(m1, m2, m3), but one of its submobjects is removed, e.g. scene.remove(m1), the list of mobjects will be edited to contain other submobjects, but not m1, e.g. it will now insert m2 and m3 to where the group once was.

Parameters¶

to_remove: The Mobject to remove.
mobject_list_name: The list of mobjects ("mobjects", "foreground_mobjects" etc) to remove from.
extract_families: Whether the mobject's families should be recursively extracted.

Returns¶

Scene: The Scene mobject with restructured Mobjects.

setup() → None¶: This is meant to be implemented by any scenes which are commonly subclassed, and have some common setup involved before the construct method is called.

should_update_mobjects() → bool¶

Returns True if the mobjects of this scene should be updated.

In particular, this checks whether

the always_update_mobjects attribute of Scene is set to True,
the Scene itself has time-based updaters attached,
any mobject in this Scene has time-based updaters attached.

This is only called when a single Wait animation is played.

tear_down() → None¶: This is meant to be implemented by any scenes which are commonly subclassed, and have some common method to be invoked before the scene ends.

property time: float¶: The time since the start of the scene.

update_mobjects(dt: float) → None¶

Begins updating all mobjects in the Scene.

Parameters¶

dt: Change in time between updates. Defaults (mostly) to 1/frames_per_second

update_self(dt: float) → None¶

Run all scene updater functions.

Among all types of update functions (mobject updaters, mesh updaters, scene updaters), scene update functions are called last.

Parameters¶

dt: Scene time since last update.

Parameters¶

duration: The run time of the animation.
stop_condition: A function without positional arguments that is evaluated every time a frame is rendered. The animation only stops when the return value of the function is truthy, or when the time specified in duration passes.
frozen_frame: If True, updater functions are not evaluated, and the animation outputs a frozen frame. If False, updater functions are called and frames are rendered as usual. If None (the default), the scene tries to determine whether or not the frame is frozen on its own.

Parameters¶

stop_condition: A function with no arguments that determines whether or not the scene should keep waiting.
max_time: The maximum wait time in seconds.

SceneFileWriter¶

负责把场景帧编码输出为视频/图片文件：分辨率、帧率、分段导出都由它落盘。

参数¶

`renderer`	—
`scene_name`	—，str

API 文档¶

class manim.SceneFileWriter(renderer: CairoRenderer | OpenGLRenderer, scene_name: str, **kwargs: Any)¶

基类：object

SceneFileWriter is the object that actually writes the animations played, into video files, using FFMPEG. This is mostly for Manim's internal use. You will rarely, if ever, have to use the methods for this class, unless tinkering with the very fabric of Manim's reality.

Attributes¶

sectionslist of Section
used to segment scene

sections_output_dirpathlib.Path
where are section videos stored

output_namestr
name of movie without extension and basis for section video names

Some useful attributes are:

"write_to_movie" (bool=False): Whether or not to write the animations into a video file.
"movie_file_extension" (str=".mp4"): The file-type extension of the outputted video.
"partial_movie_files": List of all the partial-movie files.

abort_encode_jobs(reraise_encoder_failures: bool = False) → None¶

Tear down encode jobs after an aborted or rerun render.

Seals the current job so its worker can exit (a non-daemon thread blocked on the queue would hang the process at exit), then deletes its partial file unconditionally: an aborted partial is structurally valid but truncated, so leaving it behind produces an erroneous cache hit on a later run. Sealed in-flight jobs are then drained. With reraise_encoder_failures=False (a render exception is already propagating) drain failures are logged, not raised; with True (rerun path -- no primary exception exists) the first drain failure propagates so corrupt completed partials cannot be silently ignored.

add_audio_segment(new_segment: AudioSegment, time: float | None = None, gain_to_background: float | None = None) → None¶

This method adds an audio segment from an AudioSegment type object and suitable parameters.

Parameters¶

new_segment: The audio segment to add
time: the timestamp at which the sound should be added.
gain_to_background: The gain of the segment from the background.

add_partial_movie_file(hash_animation: str | None) → None¶

Adds a new partial movie file path to scene.partial_movie_files and current section from a hash.

This method will compute the path from the hash. In addition to that it adds the new animation to the current section.

Parameters¶

hash_animation: Hash of the animation.

add_sound(sound_file: StrPath, time: float | None = None, gain: float | None = None, **kwargs: Any) → None¶

This method adds an audio segment from a sound file.

Parameters¶

sound_file: The path to the sound file.
time: The timestamp at which the audio should be added.
gain: The gain of the given audio segment.
**kwargs: This method uses add_audio_segment, so any keyword arguments used there can be referenced here.

begin_animation(allow_write: bool = False, file_path: StrPath | None = None) → None¶

Used internally by manim to stream the animation to FFMPEG for displaying or writing to a file.

Parameters¶

allow_write: Whether or not to write to a video file.

clean_cache() → None¶: Will clean the cache by removing the oldest partial_movie_files.

close_partial_movie_stream() → None¶

Close the currently opened video container.

Used internally by Manim to first flush the remaining packages in the video stream holding a partial file, and then close the corresponding container.

combine_to_movie() → None¶: Used internally by Manim to combine the separate partial movie files that make up a Scene into a single video file for that Scene.

combine_to_section_videos() → None¶: Concatenate partial movie files for each section.

create_audio_segment() → None¶: Creates an empty, silent, Audio Segment.

end_animation(allow_write: bool = False) → None¶

Internally used by Manim to stop streaming to FFMPEG gracefully.

Parameters¶

allow_write: Whether or not to write to a video file.

finish() → None¶: Finishes writing to the FFMPEG buffer or writing images to output directory. Combines the partial movie files into the whole scene. If save_last_frame is True, saves the last frame in the default image directory.

finish_last_section() → None¶: Delete current section if it is empty.

flush_cache_directory() → None¶: Delete all the cached partial movie files

get_resolution_directory() → str¶

Get the name of the resolution directory directly containing the video file.

This method gets the name of the directory that immediately contains the video file. This name is <height_in_pixels_of_video>p<frame_rate>. For example, if you are rendering an 854x480 px animation at 15fps, the name of the directory that immediately contains the video, file will be 480p15.

The file structure should look something like:

MEDIA_DIR
    |--Tex
    |--texts
    |--videos
        |--<name_of_file_containing_scene>
            |--<height_in_pixels_of_video>p<frame_rate>
                |--partial_movie_files
                |--<scene_name>.mp4
                |--<scene_name>.srt

Returns¶

str: The name of the directory.

init_audio() → None¶: Preps the writer for adding audio to the movie.

init_output_directories(scene_name: str) → None¶: Initialise output directories.

Notes¶

The directories are read from config, for example config['media_dir']. If the target directories don't already exist, they will be created.

is_already_cached(hash_invocation: str) → bool¶

Will check if a file named with hash_invocation exists.

Parameters¶

hash_invocation: The hash corresponding to an invocation to either scene.play or scene.wait.

Returns¶

bool: Whether the file exists.

join_all_encode_jobs() → None¶: Join every in-flight encode job, re-raising the first failure.

next_section(name: str, type_: str, skip_animations: bool) → None¶: Create segmentation cut here.

open_partial_movie_stream(file_path: StrPath | None = None) → None¶

Open a container holding a video stream.

This is used internally by Manim initialize the container holding the video stream of a partial movie file.

print_file_ready_message(file_path: StrPath) → None¶: Prints the "File Ready" message to STDOUT.

save_image(image: Image) → None¶

This method saves the image passed to it in the default image directory.

Parameters¶

image: The pixel array of the image to save.

write_frame(frame_or_renderer: PixelArray | OpenGLRenderer, num_frames: int = 1) → None¶

Used internally by Manim to write a frame to the FFMPEG input buffer.

Parameters¶

frame_or_renderer: Pixel array of the frame.
num_frames: The number of times to write frame.

write_subcaption_file() → None¶: Writes the subcaption file.

Section¶

场景分节：把 construct 切成带类型的片段，配合 CLI 可导出分节视频清单。

参数¶

`type_`	—，str
`video`	—，str \| None
`name`	—，str
`skip_animations`	—，bool

API 文档¶

class manim.Section(type_: str, video: str | None, name: str, skip_animations: bool)¶

基类：object

A Scene can be segmented into multiple Sections. Refer to the documentation</tutorials/output_and_config> for more info. It consists of multiple animations.

Attributes¶

type_
Can be used by a third party applications to classify different types of sections.

video
Path to video file with animations belonging to section relative to sections directory. If None, then the section will not be saved.

name
Human readable, non-unique name for this section.

skip_animations
Skip rendering the animations in this section when True.

partial_movie_files
Animations belonging to this section.

SpecialThreeDScene¶

固定相机参数的三维演示场景：低多边形默认配置，快速出 3D 效果。

继承关系¶

参数¶

`cut_axes_at_radius`	True
`camera_config`	`{'should_apply_shading': Tr…`
`three_d_axes_config`	`{'num_axis_pieces': 1, 'axi…`
`sphere_config`	{'radius': 2, 'resolution':…
`default_angled_camera_position`	{'phi': 1.2217304763960306,…
`low_quality_config`	`{'camera_config': {'should_…`

API 文档¶

class manim.SpecialThreeDScene(cut_axes_at_radius=True, camera_config={'exponential_projection': True, 'should_apply_shading': True}, three_d_axes_config={'axis_config': {'numbers_with_elongated_ticks': [0, 1, 2], 'stroke_width': 2, 'tick_frequency': 1, 'unit_size': 2}, 'num_axis_pieces': 1}, sphere_config={'radius': 2, 'resolution': (24, 48)}, default_angled_camera_position={'phi': 1.2217304763960306, 'theta': -1.9198621771937625}, low_quality_config={'camera_config': {'should_apply_shading': False}, 'sphere_config': {'resolution': (12, 24)}, 'three_d_axes_config': {'num_axis_pieces': 1}}, **kwargs)¶

基类：ThreeDScene

An extension of ThreeDScene with more settings.

It has some extra configuration for axes, spheres, and an override for low quality rendering. Further key differences are:

The camera shades applicable 3DMobjects by default, except if rendering in low quality.
Some default params for Spheres and Axes have been added.

get_axes()¶

Return a set of 3D axes.

Returns¶

ThreeDAxes: A set of 3D axes.

get_default_camera_position()¶

Returns the default_angled_camera position.

Returns¶

dict: Dictionary of phi, theta, focal_distance, and gamma.

get_sphere(**kwargs)¶

Returns a sphere with the passed keyword arguments as properties.

Parameters¶

**kwargs: Any valid parameter of Sphere or Surface.

Returns¶

Sphere: The sphere object.

set_camera_to_default_position()¶: Sets the camera to its default position.

ThreeDScene¶

三维场景：自带相机朝向设置与 3D 动画辅助（如球面移动），3D 内容入口。

继承关系¶

参数¶

`camera_class`	`<class 'manim.camera.three_…`
`ambient_camera_rotation`	None
`default_angled_camera_orientation_kwargs`	None

快速上手¶

class Demo(ThreeDScene):
    def construct(self):
        self.set_camera_orientation(phi=70 * DEGREES, theta=-45 * DEGREES)
        self.play(Create(Sphere()))

API 文档¶

class manim.ThreeDScene(camera_class=<class 'manim.camera.three_d_camera.ThreeDCamera'>, ambient_camera_rotation=None, default_angled_camera_orientation_kwargs=None, **kwargs)¶

基类：Scene

This is a Scene, with special configurations and properties that make it suitable for Three Dimensional Scenes.

add_fixed_in_frame_mobjects(*mobjects: Mobject)¶

This method is used to prevent the rotation and movement of mobjects as the camera moves around. The mobject is essentially overlaid, and is not impacted by the camera's movement in any way.

Parameters¶

*mobjects: The Mobjects whose orientation must be fixed.

add_fixed_orientation_mobjects(*mobjects: Mobject, **kwargs)¶

This method is used to prevent the rotation and tilting of mobjects as the camera moves around. The mobject can still move in the x,y,z directions, but will always be at the angle (relative to the camera) that it was at when it was passed through this method.)

Parameters¶

*mobjects

The Mobject(s) whose orientation must be fixed.

**kwargs

Some valid kwargs are: use_static_center_func : bool center_func : function

begin_3dillusion_camera_rotation(rate: float = 1, origin_phi: float | None = None, origin_theta: float | None = None)¶

This method creates a 3D camera rotation illusion around the current camera orientation.

Parameters¶

rate: The rate at which the camera rotation illusion should operate.
origin_phi: The polar angle the camera should move around. Defaults to the current phi angle.
origin_theta: The azimutal angle the camera should move around. Defaults to the current theta angle.

begin_ambient_camera_rotation(rate: float = 0.02, about: str = 'theta')¶

This method begins an ambient rotation of the camera about the Z_AXIS, in the anticlockwise direction

Parameters¶

rate: The rate at which the camera should rotate about the Z_AXIS. Negative rate means clockwise rotation.
about: one of 3 options: ["theta", "phi", "gamma"]. defaults to theta.

get_moving_mobjects(*animations: Animation)¶

This method returns a list of all of the Mobjects in the Scene that are moving, that are also in the animations passed.

Parameters¶

*animations: The animations whose mobjects will be checked.

This method animates the movement of the camera to the given spherical coordinates.

Parameters¶

phi: The polar angle i.e the angle between Z_AXIS and Camera through ORIGIN in radians.
theta: The azimuthal angle i.e the angle that spins the camera around the Z_AXIS.
focal_distance: The radial focal_distance between ORIGIN and Camera.
gamma: The rotation of the camera about the vector from the ORIGIN to the Camera.
zoom: The zoom factor of the camera.
frame_center: The new center of the camera frame in cartesian coordinates.
added_anims: Any other animations to be played at the same time.

remove_fixed_in_frame_mobjects(*mobjects: Mobject)¶

This method undoes what add_fixed_in_frame_mobjects does. It allows the mobject to be affected by the movement of the camera.

Parameters¶

*mobjects: The Mobjects whose position and orientation must be unfixed.

remove_fixed_orientation_mobjects(*mobjects: Mobject)¶

This method "unfixes" the orientation of the mobjects passed, meaning they will no longer be at the same angle relative to the camera. This only makes sense if the mobject was passed through add_fixed_orientation_mobjects first.

Parameters¶

*mobjects: The Mobjects whose orientation must be unfixed.

This method sets the orientation of the camera in the scene.

Parameters¶

phi: The polar angle i.e the angle between Z_AXIS and Camera through ORIGIN in radians.
theta: The azimuthal angle i.e the angle that spins the camera around the Z_AXIS.
focal_distance: The focal_distance of the Camera.
gamma: The rotation of the camera about the vector from the ORIGIN to the Camera.
zoom: The zoom factor of the scene.
frame_center: The new center of the camera frame in cartesian coordinates.

set_to_default_angled_camera_orientation(**kwargs)¶

This method sets the default_angled_camera_orientation to the keyword arguments passed, and sets the camera to that orientation.

Parameters¶

**kwargs: Some recognised kwargs are phi, theta, focal_distance, gamma, which have the same meaning as the parameters in set_camera_orientation.

stop_3dillusion_camera_rotation()¶: This method stops all illusion camera rotations.

stop_ambient_camera_rotation(about='theta')¶: This method stops all ambient camera rotation.

VectorScene¶

向量演示场景：坐标系 + 向量工具方法，线性代数课场景。

继承关系¶

参数¶

API 文档¶

class manim.VectorScene(basis_vector_stroke_width: float = 6.0, **kwargs: Any)¶

基类：Scene

Adds a pair of Axes to the Scene.

Parameters¶

animate: Whether or not to animate the addition of the axes through Create.
color: The color of the axes. Defaults to WHITE.

add_plane(animate: bool = False, **kwargs: Any) → NumberPlane¶

Adds a NumberPlane object to the background.

Parameters¶

animate: Whether or not to animate the addition of the plane via Create.
**kwargs: Any valid keyword arguments accepted by NumberPlane.

Returns¶

NumberPlane: The NumberPlane object.

add_vector(vector: Arrow | Vector3DLike, color: ParsableManimColor | Iterable[ParsableManimColor] = ManimColor('#FFFF00'), animate: bool = True, **kwargs: Any) → Arrow¶

Returns the Vector after adding it to the Plane.

Parameters¶

vector: It can be a pre-made graphical vector, or the coordinates of one.
color: The string of the hex color of the vector. This is only taken into consideration if 'vector' is not an Arrow. Defaults to YELLOW.
animate: Whether or not to animate the addition of the vector by using GrowArrow
**kwargs: Any valid keyword argument of Arrow. These are only considered if vector is not an Arrow.

Returns¶

Arrow: The arrow representing the vector.

coords_to_vector(vector: Vector2DLike, coords_start: Point3DLike = array([2., 2., 0.]), clean_up: bool = True) → None¶

This method writes the vector as a column matrix (henceforth called the label), takes the values in it one by one, and form the corresponding lines that make up the x and y components of the vector. Then, an Vector() based vector is created between the lines on the Screen.

Parameters¶

vector: The vector to show.
coords_start: The starting point of the location of the label of the vector that shows it numerically. Defaults to 2 * RIGHT + 2 * UP or (2,2)
clean_up: Whether or not to remove whatever this method did after it's done.

get_basis_vector_labels(**kwargs: Any) → VGroup¶

Returns naming labels for the basis vectors.

Parameters¶

**kwargs

Any valid keyword arguments of get_vector_label:: vector, label (str,MathTex) at_tip (bool=False), direction (str="left"), rotate (bool), color (str), label_scale_factor=VECTOR_LABEL_SCALE_FACTOR (int, float),

Returns a VGroup of the Basis Vectors (1,0) and (0,1)

Parameters¶

i_hat_color: The hex colour to use for the basis vector in the x direction
j_hat_color: The hex colour to use for the basis vector in the y direction

Returns¶

VGroup: VGroup of the Vector Mobjects representing the basis vectors.

get_vector(numerical_vector: Vector3DLike, **kwargs: Any) → Arrow¶

Returns an arrow on the Plane given an input numerical vector.

Parameters¶

numerical_vector: The Vector to plot.
**kwargs: Any valid keyword argument of Arrow.

Returns¶

Arrow: The Arrow representing the Vector.

get_vector_label(vector: Vector, label: ManimTextLabel | str, at_tip: bool = False, direction: str = 'left', rotate: bool = False, color: ParsableManimColor | None = None, label_scale_factor: float = 0.8) → ManimTextLabel¶

Returns naming labels for the passed vector.

Parameters¶

vector: Vector Object for which to get the label.
at_tip: Whether or not to place the label at the tip of the vector.
direction: If the label should be on the "left" or right of the vector.
rotate: Whether or not to rotate it to align it with the vector.
color: The color to give the label.
label_scale_factor: How much to scale the label by.

Returns¶

ManimTextLabel: The rendered label mobject.

label_vector(vector: Vector, label: ManimTextLabel | str, animate: bool = True, **kwargs: Any) → ManimTextLabel¶

Shortcut method for creating, and animating the addition of a label for the vector.

Parameters¶

vector: The vector for which the label must be added.
label: The rendered label mobject or the string used to create one.
animate: Whether or not to animate the labelling w/ Write
**kwargs: Any valid keyword argument of get_vector_label

Returns¶

ManimTextLabel: The rendered label mobject.

lock_in_faded_grid(dimness: float = 0.7, axes_dimness: float = 0.5) → None¶

This method freezes the NumberPlane and Axes that were already in the background, and adds new, manipulatable ones to the foreground.

Parameters¶

dimness: The required dimness of the NumberPlane
axes_dimness: The required dimness of the Axes.

show_ghost_movement(vector: Arrow | Vector2DLike | Vector3DLike) → None¶

This method plays an animation that partially shows the entire plane moving in the direction of a particular vector. This is useful when you wish to convey the idea of mentally moving the entire plane in a direction, without actually moving the plane.

Parameters¶

vector: The vector which indicates the direction of movement.

vector_to_coords(vector: Vector3DLike, integer_labels: bool = True, clean_up: bool = True) → tuple[Matrix, Line, Line]¶

This method displays vector as a Vector() based vector, and then shows the corresponding lines that make up the x and y components of the vector. Then, a column matrix (henceforth called the label) is created near the head of the Vector.

Parameters¶

vector: The vector to show.
integer_labels: Whether or not to round the value displayed. in the vector's label to the nearest integer
clean_up: Whether or not to remove whatever this method did after it's done.

write_vector_coordinates(vector: Vector, **kwargs: Any) → Matrix¶

Returns a column matrix indicating the vector coordinates, after writing them to the screen.

Parameters¶

vector: The arrow representing the vector.
**kwargs: Any valid keyword arguments of coordinate_label():

Returns¶

Matrix: The column matrix representing the vector.

ZoomedScene¶

局部放大场景：自带放大镜窗口，主画面 + 缩放插页同时呈现细节。

继承关系¶

参数¶

`camera_class`	`<class 'manim.camera.multi_…，type[Camera]`
`zoomed_display_height`	3，float
`zoomed_display_width`	3，float
`zoomed_display_center`	None，Point3DLike \| None
`zoomed_display_corner`	array([1., 1., 0.])，Vector3D
`zoomed_display_corner_buff`	0.5，float
`zoomed_camera_config`	`{'default_frame_stroke_widt…，dict[str, Any]`
`zoomed_camera_image_mobject_config`	{}，dict[str, Any]
`zoomed_camera_frame_starting_position`	array([0., 0., 0.])，Point3DLike
`zoom_factor`	0.15，float
`image_frame_stroke_width`	3，float
`zoom_activated`	False，bool

API 文档¶

class manim.ZoomedScene(camera_class: type[Camera] = <class 'manim.camera.multi_camera.MultiCamera'>, zoomed_display_height: float = 3, zoomed_display_width: float = 3, zoomed_display_center: Point3DLike | None = None, zoomed_display_corner: Vector3D = array([1., 1., 0.]), zoomed_display_corner_buff: float = 0.5, zoomed_camera_config: dict[str, Any]={'background_opacity': 1, 'default_frame_stroke_width': 2}, zoomed_camera_image_mobject_config: dict[str, Any]={}, zoomed_camera_frame_starting_position: Point3DLike = array([0., 0., 0.]), zoom_factor: float = 0.15, image_frame_stroke_width: float = 3, zoom_activated: bool = False, **kwargs: Any)¶

基类：MovingCameraScene

This is a Scene with special configurations made for when a particular part of the scene must be zoomed in on and displayed separately.

activate_zooming(animate: bool = False) → None¶

This method is used to activate the zooming for the zoomed_camera.

Parameters¶

animate: Whether or not to animate the activation of the zoomed camera.

get_zoom_factor() → float¶

Returns the Zoom factor of the Zoomed camera.

Defined as the ratio between the height of the zoomed camera and the height of the zoomed mini display.

Returns¶

float: The zoom factor.

get_zoom_in_animation(run_time: float = 2, **kwargs: Any) → ApplyMethod¶

Returns the animation of camera zooming in.

Parameters¶

run_time: The run_time of the animation of the camera zooming in.
**kwargs: Any valid keyword arguments of ApplyMethod()

Returns¶

ApplyMethod: The animation of the camera zooming in.

get_zoomed_display_pop_out_animation(**kwargs: Any) → ApplyMethod¶

This is the animation of the popping out of the mini-display that shows the content of the zoomed camera.

Returns¶

ApplyMethod: The Animation of the Zoomed Display popping out.

setup() → None¶: This method is used internally by Manim to setup the scene for proper use.

API 参考：场景（scene）¶

DefaultSectionType¶

API 文档¶

Examples¶

LinearTransformationScene¶

继承关系¶

参数¶

快速上手¶

API 文档¶

Parameters¶

Examples¶

Parameters¶

Parameters¶

Parameters¶

Parameters¶

Parameters¶

Returns¶

Parameters¶

Returns¶

Parameters¶

Parameters¶

Returns¶

Parameters¶

Returns¶

Parameters¶

Parameters¶

Parameters¶

Parameters¶

Parameters¶

Parameters¶

Parameters¶

Parameters¶

Returns¶

Parameters¶

Returns¶

Returns¶

Parameters¶

Parameters¶

Returns¶

Parameters¶

Returns¶

Parameters¶

Returns¶

MovingCameraScene¶

继承关系¶

参数¶

快速上手¶

API 文档¶

Parameters¶

Scene¶

参数¶

快速上手¶

API 文档¶

Examples¶

Parameters¶

Returns¶

Parameters¶

Returns¶

Parameters¶

Returns¶

Parameters¶

Examples¶

Parameters¶

Examples¶

Parameters¶

See also¶

Parameters¶

Returns¶

Parameters¶

Returns¶

Returns¶

Parameters¶

Returns¶

Parameters¶

Returns¶

Notes¶

Examples¶

See Also¶

Parameters¶

Returns¶