search

higlass / higlass-docker

Builds a docker container wrapping higlass-server and higlass-client in nginx
MIT License
32 stars 14 forks source link

higlass-server/manage.py ingest_tileset doesn't work #180

Open JasmineCCQ opened 2 years ago

JasmineCCQ commented 2 years ago

sudo docker run -p 11028:80 -p 11029:11029 -v ~/hg-data:/data -it --device=/dev/fuse --privileged higlass/higlass-docker /bin/bash image root@8719e80cb68c:/data# python /home/higlass/projects/higlass-server/manage.py ingest_tileset --filename /data/output.mcool --filetype cooler --datatype matrix image

when I used the commands above, it appeared this error. Thanks for any suggestion.

pkerpedjiev commented 2 years ago

Oh, can you run the container without the /bin/bash command and then open an interactive terminal using the docker exec command?

sudo docker run -p 11028:80 -p 11029:11029 -v ~/hg-data:/data --name higlass-container --device=/dev/fuse --privileged higlass/higlass-docker

And then

sudo docker exec -it higlass-container /bin/bash

If you do this, the docker run command will execute the default Dockerfile entrypoint which will set up the necessary directories and start the necessary services.

JasmineCCQ commented 2 years ago

sudo docker run -p 11028:80 -p 11029:11029 -v ~/hg-data:/data --device=/dev/fuse --privileged higlass/higlass-docker sudo docker start 66585d19b154 sudo docker exec -it 66585d19b154 /bin/bash image pip install werkzeug==2.0.0 pip install jupyter higlass-python jupyter nbextension install --py --sys-prefix --symlink higlass jupyter nbextension enable --py --sys-prefix higlass jupyter notebook --allow-root --ip='0.0.0.0' --port=11029 --no-browser image

When I modify the commands as shown above, it appears the error "Tileset info not found." I don't know why. Thanks for any suggestion!

pkerpedjiev commented 2 years ago

Is the cooler file path on your local directory or in your docker container?

With Jupyter notebook, the path should be the directory on the machine running the Jupyter notebook not the Docker container. This is because higlass-python within the Jupyter notebook starts its own higlass server.

JasmineCCQ commented 2 years ago

Thanks for your reply! My cooler file path is in my docker container. The Jupyter notebook i use is in the docker container where the higlass environment is installed. However, when i installed all the related softwares and python modules in the same conda virtual environment, and try to visualize the .mcool file, I got the error as the following. image The following is when I try to visualize the example in the higlass 'Python & Jupyter' document, I got the same error. So it may be due to my faulty process of installation, did i neglect to install anything or something? image

pkerpedjiev commented 2 years ago

Sorry for the late response but it sounds like the Jupyter extension wasn't installed properly?

Just to be sure, did you follow the instructions here exactly?