smlmlp.image_smlm module

smlmlp.image_smlm(xx, yy, image_sigma=None, weight=1.0, x_shape=None, y_shape=None, shape=None, *, pixel_sr_nm=15.0, x0=None, y0=None, crop_sigma=4.0, max_radius_pixels=0, cuda=False, parallel=False)[source]

Render localizations as an integrated Gaussian SMLM image.

Parameters:

  • xx (array-like) – Localization coordinates in nm.

  • yy (array-like) – Localization coordinates in nm.

  • image_sigma (float or array-like, optional) – Gaussian sigma in nm. Missing values default to the render pixel size.

  • weight (float or array-like, optional) – Integrated weight of each localization. Default is one count per point.

  • x_shape (int or tuple, optional) – Output image dimensions. shape is (y, x) and takes precedence. Without a shape, the image is tightly fitted around finite coordinates.

  • y_shape (int or tuple, optional) – Output image dimensions. shape is (y, x) and takes precedence. Without a shape, the image is tightly fitted around finite coordinates.

  • shape (int or tuple, optional) – Output image dimensions. shape is (y, x) and takes precedence. Without a shape, the image is tightly fitted around finite coordinates.

  • pixel_sr_nm (float, optional) – Render pixel size in nm.

  • x0 (float, optional) – Coordinate of the pixel-center origin in nm. Defaults to zero when an explicit shape is supplied, otherwise to the fitted lower bound.

  • y0 (float, optional) – Coordinate of the pixel-center origin in nm. Defaults to zero when an explicit shape is supplied, otherwise to the fitted lower bound.

  • crop_sigma (float, optional) – Gaussian support radius in sigma units.

  • max_radius_pixels (int, optional) – Optional cap for very large Gaussian radii. Zero disables the cap.

  • cuda (bool, optional) – Accepted for analysis API consistency. CPU rendering is deterministic; overlapping Gaussian writes are intentionally not parallelized.

  • parallel (bool, optional) – Accepted for analysis API consistency. CPU rendering is deterministic; overlapping Gaussian writes are intentionally not parallelized.

Returns:

  • image (ndarray) – float32 image with shape (y, x).

  • info (dict) – Rendering metadata.