hydra-jupyter is a lightweight Python interface for controlling Hydra visuals directly from Jupyter notebooks. It enables sending NumPy arrays as textures to Hydra, writing code snippets with cell magics, and controlling Hydra sessions without writing JavaScript manually.
Clone this repository and install in editable mode:
git clone https://github.com/lauraporta/hydra-jupyter.git
cd hydra-jupyter
pip install .from hydra_jupyter import HydraSession
session = HydraSession()The default hydra scrpt will be src(s0).out(o0);.
import numpy as np
# Example: simple gradient
gradient = np.tile(np.linspace(0, 1, 256), (256, 1))
session.send_array(gradient, source_id=0) # This will become your s0First, load the extension:
%load_ext hydra_jupyter.magicThen use the magic:
%%hydra
src(s0)
.kaleid(5)
.rotate(0.1)
.out(o0)def make_3d_pulsing_gradient(size=256, n_frames=30):
x = np.linspace(-1, 1, size)
y = np.linspace(-1, 1, size)
xx, yy = np.meshgrid(x, y)
radius = np.sqrt(xx**2 + yy**2)
stack = []
for i in range(n_frames):
scale = 1 + 0.5 * np.sin(2 * np.pi * i / n_frames)
pulse = np.clip(1 - radius * scale, 0, 1)
stack.append(pulse)
return np.stack(stack)
frames = make_3d_pulsing_gradient()
session.send_array(frames, frame_rate=15, source_id=0)- Send 2D or 3D NumPy arrays as textures
- Reuse persistent Hydra sessions
- Write Hydra code with
%%hydracell magic
Contributions are welcome! If you have suggestions for improvements or new features, please open an issue or submit a pull request.