View source for Data compaction

{{One source|date=June 2024}}

In [[telecommunications]], '''data compaction''' is the reduction of the number of [[data element]]s, [[Bandwidth (computing)|bandwidth]], cost, and time for the generation, [[transmission (telecommunications)|transmission]], and [[computer storage|storage of data]] without loss of [[information]] by eliminating unnecessary [[redundancy (information theory)|redundancy]], removing irrelevancy, or using special coding. 

Examples of data compaction methods are the use of fixed-[[Engineering tolerance|tolerance]] bands, variable-tolerance bands, [[slope]]-keypoints, sample changes, curve patterns, [[curve fitting]], variable-[[precision (computer science)|precision]] coding, [[frequency]] analysis, and probability analysis. 

Simply squeezing noncompacted data into a smaller space, for example by increasing packing density by transferring images from [[newsprint]] to [[microfilm]] or by transferring data on [[punched card]]s onto [[magnetic tape]], is not data compaction.{{Explain|date=November 2019|reason=What is the right name for such squeezing, if not data compaction?}}

== Everyday examples ==
The use of [[acronym]]s in [[texting]] is an everyday example.  The number of bits required to transmit and store "[[WYSIWYG]]" (What You See Is What You Get) is reduced from its expanded equivalent (7 characters vs 28).  The representation of [[Mersenne prime]]s is another example.  The largest known {{as of|2013|2|lc=on}} is over 17 million digits long but it is represented as ''M''<sub>57885161</sub> in a much more compacted form.

== See also ==

* [[brevity code]]
* [[commercial code (communications)]]
* [[data compression]]

==References==
{{FS1037C}}

[[Category:Telecommunications techniques]]
[[Category:Data compression]]


{{telecomm-stub}}