napari-broadcastable-points

Points layer that can broadcast over arbitrary dimensions. Available here as a workaround until something more complete is implemented in napari core (https://github.com/napari/napari/issues/2343).

Warning! This is likely to be very brittle for all the reasons discussed in the napari issue - so while it is useful it should also be used with caution. So don't muck around much with the viewer dims after creating, because who knows what will happen.

Installation

pip install napari-broadcastable-points

Usage

Here is an example where we have an image sequence of (TPCZYX) and we broadcast points over the C and Z axes.

import napari
from napari_broadcastable_points import BroadcastablePoints
import numpy as np

v = napari.Viewer()

# create a fake timelapse
T = 5
P = 4
C = 3
Z = 2
Y, X = 512, 512
images = np.zeros([T, P, C, Z, Y, X])
v.add_image(images)

# Add the relevant components of points data
# fmt: off
dat = np.array([
       # T,  P,       Y,              X
       [ 0,  0., 340.25071184, 284.13186557],
       [ 0,  0., 312.66551847, 309.95630191],
       [ 0,  0., 240.76794003, 266.81775485],
       [ 0,  0., 240.47448053, 239.81948049],
       [ 0,  1., 261.60356481, 260.36164576],
       [ 0,  1., 309.43746393, 215.16888217],
       [ 0,  1., 371.06395974, 235.12412843]])
# fmt: on

v.dims.axis_labels = ('T', 'P', 'C', 'Z', 'Y', 'X')

points = BroadcastablePoints(dat, broadcast_dims = (2, 3))
v.add_layer(points)
napari.run()

Creating an empty layer

You can also create an empty layer - but be sure to specify ndim otherwise you may run into an error.

points = BroadcastablePoints(None, broadcast_dims = (2, 3), ndim=6)