This project handles build and distribution of some docker images in order to extend the operations on a jupyter notebook.
Following jupyter notebook instances are built on separated images:
- machine_learning: Used for data science analysis using most popular libraries: pandas, numpy, matplotlib, etc.
- graphviz: Used for complex graph generation with graphviz library. It's configured with python api for graphviz.
- latex: Used for math documentation. It allows writing and exporting of LaTex code from inside notebook
- finance: Used for finance analysis with specific libraries.
- Ibm Watson NLP: Used for natural language processing through IBM Watson cloud api.
N.B: These are containers and by nature they are stateless instances of an image. Export files to your host machine before shutting down the container or you will lose your work.
You should have docker installed on your machine. Please follow official documentation: https://docs.docker.com/engine/installation/
You can pull the already built image from dockerhub following instructions shown at following links:
- machine_learning: https://hub.docker.com/r/lorenzomartino/machine-learning-jupyter/
- graphviz: https://hub.docker.com/r/lorenzomartino/graphviz-jupyter/
- latex: https://hub.docker.com/r/lorenzomartino/latex-jupyter/
- finance: https://hub.docker.com/r/lorenzomartino/finance-jupyter/
- Ibm Watson NLP: https://hub.docker.com/r/lorenzomartino/ibm-watson-nlp-jupyter/
Or ou can directly build and run jupyter simply executing (in case you need machine learning image)
$ docker-compose -f machine_learning/docker-compose.yml up -d
Note: You can enable widget extension inside the container running following command:
jupyter nbextension enable --py --sys-prefix widgetsnbextension
Then you should notice a log like following ones on the container side:
jupyter_1 |
jupyter_1 | Copy/paste this URL into your browser when you connect for the first time,
jupyter_1 | to login with a token:
jupyter_1 | https://0.0.0.0:8888/?token=e7dd783b5eb30462c7079b2f477eafa7024f1f981cddb99b
You should access on the host machine typing following URL on browser
https://127.0.0.1:8888/tree
It will ask you the token to be used. In the above sample is e7dd783b5eb30462c7079b2f477eafa7024f1f981cddb99b.