PMD Compression

Simple PMD compression, the hello world of masknmf

/home/runner/work/masknmf-toolbox/masknmf-toolbox/masknmf/utils/_cuda.py:10: UserWarning: You've explicitly selected to perform computations on the cpu, performance will be significantly slower
  warn(
[26-01-16 20:48:08]: Starting compression
[26-01-16 20:48:08]: sampled from the following regions: [0]
[26-01-16 20:48:08]: We are initializing on a total of 2000 frames
[26-01-16 20:48:08]: Loading data to estimate complete spatial basis
[26-01-16 20:48:08]: skipping the pruning step for frame cutoff
[26-01-16 20:48:08]: Finding spatiotemporal roughness thresholds

  0%|          | 0/250 [00:00<?, ?it/s]
  2%|▏         | 5/250 [00:00<00:05, 46.57it/s]
  4%|▍         | 10/250 [00:00<00:04, 48.09it/s]
  6%|▌         | 15/250 [00:00<00:04, 48.37it/s]
  8%|▊         | 20/250 [00:00<00:04, 48.79it/s]
 10%|█         | 25/250 [00:00<00:04, 49.18it/s]
 12%|█▏        | 30/250 [00:00<00:04, 49.09it/s]
 14%|█▍        | 35/250 [00:00<00:04, 49.02it/s]
 16%|█▌        | 40/250 [00:00<00:04, 49.20it/s]
 18%|█▊        | 45/250 [00:00<00:04, 49.24it/s]
 20%|██        | 50/250 [00:01<00:04, 49.35it/s]
 22%|██▏       | 55/250 [00:01<00:03, 49.30it/s]
 24%|██▍       | 60/250 [00:01<00:03, 49.27it/s]
 26%|██▌       | 65/250 [00:01<00:03, 49.34it/s]
 28%|██▊       | 71/250 [00:01<00:03, 49.52it/s]
 31%|███       | 77/250 [00:01<00:03, 49.65it/s]
 33%|███▎      | 82/250 [00:01<00:03, 49.69it/s]
 35%|███▍      | 87/250 [00:01<00:03, 49.52it/s]
 37%|███▋      | 93/250 [00:01<00:03, 49.71it/s]
 39%|███▉      | 98/250 [00:01<00:03, 49.72it/s]
 41%|████      | 103/250 [00:02<00:02, 49.67it/s]
 43%|████▎     | 108/250 [00:02<00:02, 49.47it/s]
 45%|████▌     | 113/250 [00:02<00:02, 49.47it/s]
 47%|████▋     | 118/250 [00:02<00:02, 49.63it/s]
 49%|████▉     | 123/250 [00:02<00:02, 49.52it/s]
 51%|█████     | 128/250 [00:02<00:02, 49.46it/s]
 54%|█████▎    | 134/250 [00:02<00:02, 49.62it/s]
 56%|█████▌    | 139/250 [00:02<00:02, 49.69it/s]
 58%|█████▊    | 144/250 [00:02<00:02, 49.54it/s]
 60%|█████▉    | 149/250 [00:03<00:02, 49.53it/s]
 62%|██████▏   | 154/250 [00:03<00:01, 49.53it/s]
 64%|██████▎   | 159/250 [00:03<00:01, 49.39it/s]
 66%|██████▌   | 164/250 [00:03<00:01, 49.53it/s]
 68%|██████▊   | 169/250 [00:03<00:01, 49.58it/s]
 70%|██████▉   | 174/250 [00:03<00:01, 49.53it/s]
 72%|███████▏  | 179/250 [00:03<00:01, 49.53it/s]
 74%|███████▎  | 184/250 [00:03<00:01, 49.55it/s]
 76%|███████▌  | 189/250 [00:03<00:01, 49.60it/s]
 78%|███████▊  | 195/250 [00:03<00:01, 49.68it/s]
 80%|████████  | 200/250 [00:04<00:01, 49.56it/s]
 82%|████████▏ | 205/250 [00:04<00:00, 49.59it/s]
 84%|████████▍ | 210/250 [00:04<00:00, 49.40it/s]
 86%|████████▌ | 215/250 [00:04<00:00, 49.40it/s]
 88%|████████▊ | 220/250 [00:04<00:00, 49.28it/s]
 90%|█████████ | 225/250 [00:04<00:00, 49.19it/s]
 92%|█████████▏| 230/250 [00:04<00:00, 49.09it/s]
 94%|█████████▍| 235/250 [00:04<00:00, 47.72it/s]
 96%|█████████▌| 240/250 [00:04<00:00, 48.08it/s]
 98%|█████████▊| 245/250 [00:04<00:00, 48.52it/s]
100%|██████████| 250/250 [00:05<00:00, 48.74it/s]
100%|██████████| 250/250 [00:05<00:00, 49.27it/s]
[26-01-16 20:48:13]: Running Blockwise Decompositions
[26-01-16 20:48:13]: Constructed U matrix. Rank of U is 192
[26-01-16 20:48:13]: PMD Objected constructed
/opt/hostedtoolcache/Python/3.12.12/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()