3D cell shape analysis using geometric deep learning on point clouds
Project description
Cellshape-cloud is an easy-to-use tool to analyse the shapes of cells using deep learning and, in particular, graph-neural networks. The tool provides the ability to train popular graph-based autoencoders on point cloud data of 2D and 3D single cell masks as well as providing pre-trained networks for inference.
To install
pip install cellshape-cloud
Usage
Basic Usage
import torch
from cellshape_cloud import CloudAutoEncoder
model = CloudAutoEncoder(num_features=128,
k=20,
encoder_type="dgcnn",
decoder_type="foldingnet")
points = torch.randn(1, 2048, 3)
recon, features = model(points)
To train an autoencoder on a set of point clouds created using cellshape-helper:
import torch
from torch.utils.data import DataLoader
import cellshape_cloud as cloud
from cellshape_cloud.vendor.chamfer_distance import ChamferLoss
input_dir = "path/to/pointcloud/files/"
batch_size = 16
learning_rate = 0.0001
num_epochs = 1
output_dir = "path/to/save/output/"
model = cloud.CloudAutoEncoder(num_features=128,
k=20,
encoder_type="dgcnn",
decoder_type="foldingnet")
dataset = cloud.PointCloudDataset(input_dir)
dataloader = DataLoader(dataset, batch_size=batch_size, shuffle=True)
criterion = ChamferLoss()
optimizer = torch.optim.Adam(
model.parameters(),
lr=learning_rate * 16 / batch_size,
betas=(0.9, 0.999),
weight_decay=1e-6,
)
cloud.train(model, dataloader, num_epochs, criterion, optimizer, output_dir)
Parameters
num_features
: int.
The size of the latent space of the autoencoder.k
: int.
The number of neightbours to use in the k-nearest-neighbours graph construction.encoder_type
: str.
The type of encoder: 'foldingnet' or 'dgcnn'decoder_type
: str.
The type of decoder: 'foldingnet' or 'dgcnn'
For developers
- Fork the repository
- Clone your fork
git clone https://github.com/USERNAME/cellshape-cloud
- Install an editable version (
-e
) with the development requirements (dev
)
cd cellshape-cloud
pip install -e .[dev]
- To install pre-commit hooks to ensure formatting is correct:
pre-commit install
- To release a new version:
Firstly, update the version with bump2version (bump2version patch
,
bump2version minor
or bump2version major
). This will increment the
package version (to a release candidate - e.g. 0.0.1rc0
) and tag the
commit. Push this tag to GitHub to run the deployment workflow:
git push --follow-tags
Once the release candidate has been tested, the release version can be created with:
bump2version release
References
[1] An Tao, 'Unsupervised Point Cloud Reconstruction for Classific Feature Learning', GitHub Repo, 2020
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for cellshape-cloud-0.0.14rc0.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | a9083b1218cc8baa8dab7f6ce0c79d0b09171197b7f22a1314ac864b82f10b37 |
|
MD5 | 05659ad2d889a5b183c66e116eb4d5ff |
|
BLAKE2b-256 | 5c0af31ff777762fcff3a940621dcfd3e86cbed9d528a48b7e1db289f87820af |
Hashes for cellshape_cloud-0.0.14rc0-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | efad3f16443a4eed41cd101c24f41094f7643cee15a5386fe498514cef910bc8 |
|
MD5 | d09fb7f694733d1fde76bab8ac1523d1 |
|
BLAKE2b-256 | e93787f5120fa671c8c561ece895d474a02a34ef26ce869e15fdcc6e8e839025 |