The APL language and statistical computations
K. W. Smillie · ITNOW · 1969
APL, for A Programming Language, was developed by K. E. Iverson of the IBM Thomas J. Watson Research Center who gave a thorough treatment of the language in Iverson [1]. Since then, the concepts of the language and its notation have been continually developed and refined by Iverson and his colleagues who have implemented a version of the language on several IBM computers including some of the larger System/360 models. A complete description of the language as presently implemented on the System/360 is given in Falkoff and Iverson [2], while a summary appears in Falkoff and Iverson [3]. The language has been implemented in an interpretive interactive mode with all input and output through an IBM 2741 communications terminal similar to an electric typewriter. Each user has available to him an arbitrary number of blocks, called workspaces, of memory, each consisting of 31,872 bytes, and collectively referred to as a library. While programs, called functions in APL, and data may be transferred from one workspace to another, even though they may be in different libraries, during any computation all data, the functions which operate on these data, and all intermediate and final results which are produced by these operations must be contained in the one workspace which has currently been made active.