![]() ![]() open Anaconda navigation (based on your OS) Anaconda Jupyter Notebook MAC,Windows, and Linux After installing anaconda (version 3.x), you can follow the steps below to start jupyter notebook. Though there are multiple ways to install and run jupyter notebook, Anaconda is the preferred way to run jupyter. You can use jupyter notebook to write, execute and document python code. A notebook document can contain code, text, equations and/or visualizations. Jupyter notebook is a web application to create and share documents. Please check the user guide for more details: One of those tools is jupyter notebook that will be used to run python code. ![]() Open Anaconda navigation to check the tools that it provides: ![]() General details of how to install Anaconda, can be found here Anaconda - Usage on the middle of the page click on Linux follow the instruction on the page Linux download "Python 3.7 version" 32-bit or 64-bit. on the middle of the page click on Windows follow the instructions on either of these pages: on the middle of the page click on macOS To overcome the issue of outdated and/or multiple versions of python, we will use anaconda to install python version 3.x, and to run jupyter notebooks (explained later). Python is already installed on MAC and Linux machines. Now open firefox, or whatever web browser you are using and enter the following address: You will be asked the password you setup at the begining.Ĭlose your local terminal when your are done.(Acknowledgment: This is a slightly modified version of instructions by Faras Sadek of SEAS Computing) Python Leave the terminal open and running, this is the client end of the tunnel. You should be asked your password, possibly 2 times, and then the teminal should print a bunch of lines ending with debug1: client_input_global_request: rtype want_reply 0 Replace PORT by the port number you picked earlier ssh -v -N -L PORT:localhost:PORT ebi_jupyter This is the only step that you will have to rerun everytime you want to use a notebook on the cluster Prox圜ommand ssh -o tisauthentication=yes gate proxy %h 2>/dev/null Create it if it doesn't exist yet.Īdd the following lines, replacing HOSTNAME by the host name you just wrote down and USERNAME by your ebi user name Host gate Write down the host name On you local machine Edit you ssh config fileĮdit you config file ~/.ssh/config. In the example below it is running on hx-noah-09-05. JOBID USER STAT QUEUE FROM_HOST EXEC_HOST JOB_NAME SUBMIT_TIMEĥ89876 aleg RUN dungeon-rh74 ebi-cli-003 4*hx-noah-09-05 jupyter_server Apr 9 16:35 Write down this numberįind the host on which your notebook is running bjobs -w Replace the port by a random 4 digit number greater than 9000 (please don't use 9387 as this is the one I use). Then submit a bsub job to the dungeon queue so that Jupyter is always serving your notebooks bsub -R "span" -q dungeon-rh74 -J jupyter_server -M 5000 -n 4 "source ~/.bashrc & conda activate jupyter & jupyter notebook -no-browser -port=PORT" Then go to the directory where you have all your analyses/datasets cd ROOT/OF/YOUR/ANALYSIS/FOLDER These steps should only be ran a single time and then once in a while, when your job dies for whatever reasonsįirst deactivate conda environment conda deactivate Jt -t grade3 -fs 95 -altp -tfs 11 -nfs 115 -cellw 88% -TĪlways in jupyter conda environment conda install -c conda-forge numpy scipy matplotlib seaborn pandas Jt -t onedork -fs 95 -altp -tfs 11 -nfs 115 -cellw 88% -T It is also useful to install jupyter extensions conda install -c conda-forge jupyter_contrib_nbextensionsĬonda install -c conda-forge jupyter_nbextensions_configuratorĪnd customize the default CSS template with jupyter-themes conda install -c conda-forge jupyterthemes Setup a password to protect your notebook from being accessed by anyone. Install nupyter notebook from conda conda activate jupyter See Create a new python 3.6 or 3.7 conda environmentĬonda create -n jupyter python=3.7 Install Jupyter notebook in the new environment Install Jupyter notebook On the cluster Install Conda if not already done ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |