A Hitchhiker's Guide to Data Science
summary
Overview of concrete claims and tools: Anaconda ships with conda, conda-forge and Anaconda Cloud and now bundles MKL optimizations (previously paid) for faster numeric work; pandas gains SAS readers and notebook conditional formatting; Jupyter supports multiple kernels including IR and NBextensions like NBpresent. Bokeh rewrites its server on tornado/websocket and powers server apps; DataShader produces density-aware plots to avoid overplotting and integrates with Dask and Numba to scale; Dask 10–100× faster for array work and Anaconda with conda and conda-forge makes reproducible environments straightforward.
AI-written summary · about
chapters (11)
- 0:00Welcome
- 1:48Data science workflows
- 3:57Anaconda ecosystem
- 8:31Anaconda cloud
- 10:00Jupyter notebooks
- 13:11Data exploration with pandas
- 19:33Visualizing large datasets
- 25:55Scaling with Dask
- 27:40Speeding up Python
- 31:33Summary
- 31:54Q&A
related talks
38mState of Data Science & Machine Learning
43mKEYNOTE: Scaling Out PyData
36mAnaconda Ecosystem for Open Data Science
31mPython Anytime, Anywhere with Anaconda Notebooks
33mData Science & Data Visualization in Python
37mThe Python Ecosystem for Data Science: A Guided Tour
30mEmpowered by Python - A success story
28mUsing Python in a Data Hackathon