Active Learning with Jupyter Notebooks

This is a Python 3 library to read/write cells programmatically in Jupyter notebooks <https://jupyter.org/>_ which anticipates upcoming collaborative <https://groups.google.com/forum/#!topic/jupyter/r7QSObF5YSg>_ features in Jupyter.

We use this at O'Reilly Media <https://www.oreilly.com/>_ for notebooks used to manage machine learning pipelines. That is to say, machines and people collaborate on documents, implementing a "human-in-the-loop" design pattern:

• people adjust parameters for the ML pipelines
• machines write structured "logs" during ML modeling/evaluation
• people run jupyter notebook via SSH tunnel for remote access

For more info about use cases for this library and active learning in general, see the JupyterCon 2017 <https://jupytercon.com/>_ talk Humans in the loop <https://conferences.oreilly.com/jupyter/jup-ny/public/schedule/detail/60058>_

Example Usage

The following script generates a Jupyter notebook in the test.ipynb file, then runs it:

::

python test.py
jupyter notebook

Then launch the test.ipynb notebook and from the Cells menu select Run All to view results.

NB: whenever you use the put_df() function to store data as a Pandas dataframe <https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.html>_ be sure to include import pandas as pd at some earlier point in the notebook.

Dependencies and Installation

This code has dependencies on:

• nbformat <https://github.com/jupyter/nbformat>_
• pandas <https://pandas.pydata.org/>_

To install from PyPi <https://pypi.python.org/pypi/nbtransom>_:

::

pip install nbtransom

To install from this Git repo:

::

git clone https://github.com/ceteri/nbtransom.git
cd nbtransom
pip install -r requirements.txt

Kudos

@htmartin <https://github.com/htmartin>_ @esztiorm <https://github.com/esztiorm>_ @fperez <https://github.com/fperez>_ @odewahn <https://github.com/odewahn>_