Getting Started

This example builds a scene containing the Suzanne model, two cylinders, and a text label. Install FURY first using the Installation guide.

Import the current scene, actor, UI, and array-based I/O APIs:

from pathlib import Path

import numpy as np

from fury import actor, io, ui, window
from fury.data import fetch_viz_models

Download the example models once. This requires network access on the first run; subsequent runs reuse the cached files. The fetcher returns file information and the model directory:

_, model_dir = fetch_viz_models()
path_suzanne = Path(model_dir) / "suzanne.obj"

Read the mesh as NumPy arrays and create a partially transparent surface. The Suzanne model has triangle faces; a point-only file cannot be used for this surface example:

vertices, faces, colors = io.read_mesh(str(path_suzanne))
if faces is None:
    raise ValueError("The Suzanne example requires a triangle mesh.")
suzanne = actor.surface(vertices, faces, colors=colors, opacity=0.5)

Create two cylinders with explicit positions, directions, heights, and colors:

centers = np.array([[2., 0., 0.], [-2., 0., 0.]])
directions = np.array([[0., 1., 0.], [0., 1., 0.]])
heights = np.array([1., 1.5])
colors = np.array([[1., 0., 0.], [0., 0., 1.]])
cylinders = actor.cylinder(
    centers, directions=directions, colors=colors, height=heights, radii=0.2
)

Add a dynamically sized text label and all actors to the scene before creating the rendering window:

tb = ui.TextBlock2D(
    text="Hello FURY", position=(20, 20), font_size=23, dynamic_bbox=True
)
scene = window.Scene()
scene.add(suzanne)
scene.add(cylinders)
scene.add(tb)

Interactive rendering

Create the manager using keyword arguments and start the event loop. Close the window to return to Python:

showm = window.ShowManager(scene=scene, size=(1024, 720))
showm.start()

Offscreen alternative

Instead of the interactive manager and start() block above, save the scene directly to an image:

window.snapshot(scene=scene, fname="getting_started.png")

Offscreen rendering still requires a working graphics backend. See Installation for graphics requirements and headless testing guidance.