I needed an Anaconda environment for a training class, but did not want to install the tools locally. After not finding a simply, complete and working Docker example I put this project together. I am sure it is missing some things and I will be updating stuff for as long as my training goes or based on how much I use Python. Hopefully this will help others starting out that just want a simple, run-it-and-forget-it example to hit the floor running.
Note: This application has not been hardened or vulnerability tested and is not intended for production use.
- This uses continuumio/anaconda3 which I assume is latest. If things break for you then it might be that anaconda was updated. I was using Anaconda 2019.03 when I created this project.
- jupyter_notebook_config.json and requirements.txt are used in the gitignore to avoid them being uploaded for my projects. Simply rename the existing files in /srv/anaconda/conf to remove _EXAMPLE and the project should function as a basic level.
- Password for Notebook is "password" if you use the examples included. If you want to change this:
- Blank out local file /srv/anaconda/conf/jupyter_notebook_config.json
- Compose Up and visit http://localhost
- Shell into Container and run "/opt/conda/bin/jupyter notebook list"
- This will output a URL with the login Token.
- Use this Token to set a password and login.
- In your Container Shell, cp /root/.jupyter/jupyter_notebook_config.json /opt/notebooks/jupyter_notebook_config.json
- This will place the config file in the notebooks folder in your local project.
- Move this file to overwrite the local file /srv/anaconda/conf/jupyter_notebook_config.json
- Compose Down and Compose Up and Notebooks should use your new password.
- /srv/anaconda/conf/requirements.txt should list out any required packaged you need for PIP to install.
/srv
- Custom configs for Anaconda
- Contains Dockerfile for Anaconda build.
/notebooks
- Source Files for Notebooks