ADSP: Algorithms + Data Structures = Programs Episode 0: Our Favorite Algorithms
5 snips
Nov 20, 2020 AI Snips
Chapters
Transcript
Episode notes
How Transform Reduce Sparked A Friendship
- Connor and Bryce bonded after Connor's talk about C++ sequence algorithms, especially transform_reduce, at a Bay Area C++ user group and C++Now 2019.
- Bryce had written specs for many algorithms and was thrilled Connor identified their problems, sparking their friendship and collaboration.
Transform Reduce Unifies Map And Reduce
- Transform_reduce (map then reduce) composes a per-element transform with a binary reduction, enabling concise descriptions of many computations.
- Bryce and Connor highlight its power for data science and parallel implementations where transform and reduction separate concerns for efficiency.
Outer Product Produces Matrix Results
- Outer product produces a matrix (range of ranges) by applying a binary operation to every pair from two sequences, common in array languages.
- Connor notes APL, J, Julia, and Haskell support outer product and it yields multiplication tables or other tensor-like results.