PMD Compression

Simple PMD compression, the hello world of masknmf

[25-10-29 21:35:13]: Starting compression
[25-10-29 21:35:13]: sampled from the following regions: [0]
[25-10-29 21:35:13]: We are initializing on a total of 2000 frames
[25-10-29 21:35:13]: Loading data to estimate complete spatial basis
[25-10-29 21:35:13]: skipping the pruning step for frame cutoff
[25-10-29 21:35:13]: Finding spatiotemporal roughness thresholds

  0%|          | 0/250 [00:00<?, ?it/s]
  3%|▎         | 7/250 [00:00<00:03, 62.40it/s]
  6%|▌         | 14/250 [00:00<00:03, 64.17it/s]
  8%|▊         | 21/250 [00:00<00:03, 64.57it/s]
 11%|█         | 28/250 [00:00<00:03, 65.24it/s]
 14%|█▍        | 35/250 [00:00<00:03, 66.28it/s]
 17%|█▋        | 43/250 [00:00<00:03, 67.78it/s]
 20%|██        | 50/250 [00:00<00:02, 67.49it/s]
 23%|██▎       | 57/250 [00:00<00:02, 67.68it/s]
 26%|██▌       | 64/250 [00:00<00:02, 68.22it/s]
 28%|██▊       | 71/250 [00:01<00:02, 68.32it/s]
 31%|███       | 78/250 [00:01<00:02, 67.95it/s]
 34%|███▍      | 85/250 [00:01<00:02, 67.93it/s]
 37%|███▋      | 92/250 [00:01<00:02, 67.26it/s]
 40%|███▉      | 99/250 [00:01<00:02, 67.61it/s]
 42%|████▏     | 106/250 [00:01<00:02, 67.75it/s]
 45%|████▌     | 113/250 [00:01<00:02, 68.00it/s]
 48%|████▊     | 120/250 [00:01<00:01, 66.68it/s]
 51%|█████     | 127/250 [00:01<00:01, 67.04it/s]
 54%|█████▎    | 134/250 [00:01<00:01, 67.33it/s]
 56%|█████▋    | 141/250 [00:02<00:01, 67.39it/s]
 59%|█████▉    | 148/250 [00:02<00:01, 67.55it/s]
 62%|██████▏   | 155/250 [00:02<00:01, 67.66it/s]
 65%|██████▍   | 162/250 [00:02<00:01, 67.08it/s]
 68%|██████▊   | 169/250 [00:02<00:01, 66.84it/s]
 70%|███████   | 176/250 [00:02<00:01, 66.27it/s]
 73%|███████▎  | 183/250 [00:02<00:01, 65.61it/s]
 76%|███████▌  | 190/250 [00:02<00:00, 65.30it/s]
 79%|███████▉  | 197/250 [00:02<00:00, 64.73it/s]
 82%|████████▏ | 204/250 [00:03<00:00, 64.46it/s]
 84%|████████▍ | 211/250 [00:03<00:00, 64.94it/s]
 87%|████████▋ | 218/250 [00:03<00:00, 65.69it/s]
 90%|█████████ | 225/250 [00:03<00:00, 65.43it/s]
 93%|█████████▎| 232/250 [00:03<00:00, 65.14it/s]
 96%|█████████▌| 239/250 [00:03<00:00, 65.34it/s]
 98%|█████████▊| 246/250 [00:03<00:00, 65.72it/s]
100%|██████████| 250/250 [00:03<00:00, 66.51it/s]
[25-10-29 21:35:17]: Running Blockwise Decompositions
[25-10-29 21:35:17]: Constructed U matrix. Rank of U is 182
[25-10-29 21:35:17]: PMD Objected constructed
/opt/hostedtoolcache/Python/3.12.11/x64/lib/python3.12/site-packages/fastplotlib/graphics/features/_base.py:18: UserWarning: casting float64 array to float32
  warn(f"casting {array.dtype} array to float32")

# test_example = true

import masknmf
import torch
import fastplotlib as fpl
from urllib.request import urlretrieve
import tifffile


urlretrieve(
    "https://github.com/flatironinstitute/CaImAn/raw/refs/heads/main/example_movies/demoMovie.tif",
    "./demo.tif"
)

# always lazy load raw data by memmaping or other methods
data = tifffile.imread("./demo.tif")

block_sizes = [32, 32]
max_components = 20

# it's recommended to use masknmf on a machine with a GPU
# device = "cuda" if torch.cuda.is_available() else "cpu"
device = "cpu"
# number of frames used to estimate the spatial basis in PMD
num_frames_for_spatial_fit = data.shape[0]

# perform PMD
pmd_result = masknmf.compression.pmd_decomposition(
    data,
    block_sizes,
    num_frames_for_spatial_fit,
    max_components=max_components,
    device=device,
    frame_batch_size=1024
)

# get the residual
pmd_residual = masknmf.PMDResidualArray(data, pmd_result)

# view the movies, note that all these array are LAZY evaluated, allowing you to view extremely large datasets!
iw = fpl.ImageWidget(
    data=[data, pmd_result, pmd_residual],
    names=["raw", "pmd", "residual"],
    figure_kwargs={"size": (1000, 340), "shape": (1, 3)},
    cmap="gnuplot2",
)

iw.show()

# use the time slider or set the frame index programmatically
iw.current_index = {"t": 1610}

# manually set vmin-vmax to emphasize noise in raw video
# you can also adjust the vmin-vmax using the histogram tool
# reset the vmin-vmax by clicking the buttons under "ImageWidget Controls"
for image in iw.managed_graphics:
    image.vmax = 3_200

# remove toolbar to reduce clutter
for subplot in iw.figure:
    subplot.toolbar = False


# ignore the remaining lines these are just for docs generation
figure = iw.figure
if __name__ == "__main__":
    print(__doc__)
    fpl.loop.run()