NumPy & SciPy with Travis Oliphant
January 27th, 2021
49 mins 31 secs
About this Episode
Eric Anderson (@ericmander) and Travis Oliphant (@teoliphant) take a far-reaching tour through the history of the Python data community. Travis has had a hand in the creation of many open-source projects, most notably the influential libraries, NumPy and SciPy, which helped cement Python as the standard for scientific computing. Join us for the story of a fledgling community from a time “before open-source was cool,” and their lessons for today’s open-source landscape.
In this episode we discuss:
- How biomedical engineering, MRIs, and an unhappy tenure committee led to NumPy and SciPy
- Overcoming early challenges of distribution with Python
- What Travis would have done differently when he wrote NumPy
- Successfully solving the “two-option split” by adding a third option
- Community-driven open-source interacting with company-backed open-source
Links:
- NumPy
- SciPy
- Anaconda
- Quansight
- Conda
- Matplotlib
- Enthought
- TensorFlow
- PyTorch
- MXNet
- PyPi
- Jupyter
- pandas
People mentioned:
- Guido van Rossum (@gvanrossum)
- Robert Kern (Github: @rkern)
- Pearu Peterson (Github: @pearu)
- Wes McKinney (@wesmckinn)
- Charles Harris (Github: @charris)
- Francesc Alted (@francescalted)
- Fernando Perez (@fperez_org)
- Brian Granger (@ellisonbg)
Other episodes: