Index origin

From APL Wiki
Revision as of 14:40, 17 October 2019 by Miraheze>Marshall (Created page with "The index origin is the number used for the first index along each axis of an array. Many APLs allow the user to configure index origin using the system variable <...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

The index origin is the number used for the first index along each axis of an array. Many APLs allow the user to configure index origin using the system variable ⎕IO. Even array languages which do not have such a configuration parameter must make a choice of index origin; for example, J uses a non-configurable index origin of 0. The English language uses an index origin of one: thus the element with index ⎕IO in APL is referred to in English as the "first" element.

The index origin affects both the handling of index arguments to functions (and other functionality like axis specification) and their results. For example, Iota (or "Index Generator") creates results which begin with ⎕IO.

      ⎕IO←0
      ⍳4
0 1 2 3
      ⎕IO←1
      ⍳4
1 2 3 4

Any number could consistently be used for the index origin, but almost all APLs restrict the choice to 0 or 1. ngn/apl supports ⎕IO but only allows it to have the value 0. Attempting to assign an unsupported value to ⎕IO typically results in a DOMAIN ERROR.

The choice of which numbers are indices is subjective and must be made by the language designer. In some cases there is no clear decision: Interval Index and Occurrence Count each have natural definitions that depend on index origin, and natural definitions that do not.

External links

Wikipedia

Is Index Origin 0 a Hindrance?