Try it out: the current url is at https://uncurl.cs.washington.edu. Alternatively, try https://uncurl-test.cs.washington.edu.
Requirements: see requirements.txt. To install requirements, run pip install -r requirements.txt
To install: run pip install -e . - this is necessary to use the uncurl_app_split_seq script (see user_guide.md).
To deploy locally: run python run.py
To deploy on a server (requires redis): run sh start-gunicorn.sh
Main code is located in uncurl_app/
By default, data is stored at /tmp/uncurl/.
To build a new image from the repository's root directory:
docker build . -t uncurl-app
To run the server:
docker run -p 6379:6379 -p <port>:8888 uncurl-app
This exposes the given port, and then the uncurl-app website can be visited in the browser at http://your-ip-address:port. To stop the server, run sudo killall gunicorn in another terminal.
Alternatively, we have built reasonably up-to-date images at ayuezhang27/uncurl-app. To run the server using these images (does not require cloning this repository):
docker pull ayuezhang27/uncurl-app
docker run -p 8888:8888 -p 6379:6379 ayuezhang27/uncurl-app
When deploying on AWS, make sure to allow HTTP requests to and from the selected port in the security group.
Pushing to docker-hub:
docker build . -t uncurl-app
docker tag uncurl-app ayuezhang27/uncurl-app
docker push ayuezhang27/uncurl-app
python test/test_app.py
python test/test_frontend.py (requires selenium and the Firefox webdriver from https://github.com/mozilla/geckodriver/releases)
See http://uncurl.cs.washington.edu/help
In order to change the file upload size limits when running locally, set the MAX_CONTENT_LENGTH environment variable to the desired value (in bytes).
Most of the features of uncurl-app are provided through the uncurl_analysis package. The SCAnalysis class represents the data analysis.
Cell type databases include cellmarker_python and cellmesh.
loading gif from http://www.ajaxload.info/