The Story in the Notebook: Exploratory Data Science using a Literate Programming Tool
summary
21 data scientists interviewed at JupyterCon 2017 and a survey of 45 users show that Jupyter notebooks drive exploration but rarely produce prose narratives: only six participants use formatted annotations during active work and three use them to record thought processes. Users iterate by creating many small cells then expand–reduce curation, with 11 people reporting deleting failed analyses; seven run notebooks as production pipelines and four preserved full experiment records, revealing a tradeoff between clean stories and complete history.
AI-written summary · about
chapters (9)
- 0:00Introduction to literate programming
- 2:57History of literate programming
- 4:42Jupyter notebooks in data science
- 7:11Narrative use in notebooks
- 9:15Annotations during exploration
- 10:38Iteration and cell curation
- 13:23Exploratory record challenges
- 15:07Summary and future directions
- 15:29Q&A session
related talks
17mNotes on Notebooks: Is Jupyter the Bringer of Jollity?
49mBest tools for Jupyter notebooks
25mOvercoming 5 Hurdles to Using Jupyter Notebooks for Data Science, by the JetBrai...
49mExperimenting with Large Data Volumes Using Notebooks
28mUnlock the Full Potential of Jupyter Notebooks
33mSignificant Roadblocks to Usefulness for Jupyter Notebooks and a Recipe to Fix them
25mThe First Notebook War
29mOh no! I think my project is outgrowing my Jupyter notebook. How do I survive?