理解向量化编程Vectorized Implementation

• Vectorization

  “Vectorization” is the process of rewriting a loop so that instead of processing a single element of an array N times.

• Array Programming

  Array programming refers to solutions which allow the application of operations to an entire set of values at once.

  Modern programming languages that support array programming (also known as vector or multidimensional languages) have been engineered specifically to generalize operations on scalars to apply transparently to vectors, matrics, and higher-dimensional arrays.

  Important people for Array Programming: Kenneth E. Iverson、Arthur Whitney

  APL wiki

• Kenneth E. Iverson

  Kenneth E. Iverson (1920.12.17-2004.10.19) was a Canadian computer scientist noted for the development of the programming language APL(named after the book A Programming Language).

• Arthur Whitney

  Arthur Whitney(1957.10.20-) is a Canadian computer scientist most notable for developing three programming languages inspired by APL: A+, k, q.

  Whitney is recognized as having an “enduring and significant influence on APL”;

  He also wrote the intial prototype of J;

  In 1988, Whitney began working at Morgan Stanley developing financial applicaitons; At Morgan Stanley, Whitney developed A+ to facilitate migrating APL applications from IBM mainframe computers to a network of Sun Microsystems workstations;

  In 1993, Whitney left Morgan Stanley and co-founded Kx Systems with Janet Lustgarten, to commercialize his k programming language; Initially, Kx Systems had an exclusive contract with Swiss global financial services firm UBS to provide them with the k language; At the outset of the contract Whitney developed the kdb database built on k;

  In 1998, the contract with UBS expired and the firm launched the kdb+ database; as part of kdb+, Whitney developed a new language named q that operates with k and uses English keywords;

  In 1999, the firm reached a marketing agreement with Northern Ireland based firm, First Derivatives;

  In 2003, Kx Systems released q, a new vector language that built upon k and kdb+ database developed by Whitney;

  In 2014, First Derivatives purchased a 65 percent share of Kx Systems, which began a trend in offering Kdb+ to other data-driven industries;

  In 2018, First Derivatives bought out Whitney and Lustgarten’s minority shares of Kx Systems;

  Whitney and Lustgarten then founded Shakti;

  The Shakti platform has a small memory footprint, and allows for fast deployment and processing of distributed elastic workloads. It can work with all kinds of datasets, including numerical, temporal and text data, whether structured or not.

  As of 2019 Kdb+ is availble as an on-demand solution via the Amazon Web Services Marketplace;

  A+

  K

  Q

• References

1. Vectorization Implementation in Machine Learning
2. StackOverflow: What is “vectorization”?