Align2D

class quantem.widget.Align2D(**kwargs: Any)

Bases: AnyWidget

Interactive alignment of two 2D images.

Parameters:

• image_a (array_like) – First 2D image (reference, stays fixed).

• image_b (array_like) – Second 2D image (draggable).

• title (str, optional) – Title displayed above the viewer.

• label_a (str, default "Image A") – Label for the first image.

• label_b (str, default "Image B") – Label for the second image.

• cmap (str, default "gray") – Colormap name.

• opacity (float, default 0.5) – Blend ratio (0 = only A, 1 = only B).

• padding (float, default 0.2) – Fractional padding on each side. Adjustable from the frontend.

• pixel_size (float, default 0.0) – Pixel size in Å for scale bar (0 = uncalibrated).

• canvas_size (int, default 300) – Initial canvas size in CSS pixels for each column.

• auto_align (bool, default True) – Automatically compute initial alignment via cross-correlation.

• max_shift (float, default 0.0) – Maximum allowed shift in pixels (0 = unlimited, constrained by padding).

• rotation (float, default 0.0) – Initial rotation angle of image B in degrees.

• hist_source (str, default "a") – Which image to show in the histogram (“a” or “b”).

Examples

>>> import numpy as np
>>> from quantem.widget import Align2D
>>> a = np.random.rand(64, 64).astype(np.float32)
>>> b = np.random.rand(64, 64).astype(np.float32)
>>> Align2D(a, b, title="Alignment")

Key Methods

Align2D.set_images(image_a, image_b, auto_align=True)

Replace both images. Preserves display settings, recomputes alignment.

Align2D.reset_alignment()

State Persistence

Align2D.state_dict()

Align2D.save(path: str)

Align2D.load_state_dict(state)

Align2D.summary()