SAP Sorting and Search algorithms

[Former Member]

Greetings,

I've been looking around but could not find a good answer regarding what kind of algorithms SAP uses for Sorting and searching.

My guess is that SAP uses quick sort or radix sorts perhaps for sorting and linear search/brute force for reads. Any input on this would be appreciated.

[former_member192616]

Hi,

are you talking about internal tables?

> My guess is that SAP uses quick sort or radix sorts perhaps for sorting

i learnt it is something "very similar to quick sort"

> and linear search/brute force for reads. Any input on this would be appreciated.

linear, binary or hash search alogrithms are used (depends e.g. on table type)

Kind regards,

Hermann

[Former Member]

Hi,

Indeed, I'm talking about internal tables.

If anyone can add more comments to confirm the algorithms used (specially for the sorting part) with a reliable source where the information came from, I'm good to go.

[Former Member]

This message was moderated.

[Former Member]

Hi, Search/Read part is already answered. Thanks.

To fully answer my concern, I require confirmation (with a source,such as an article, book citation, book, etc.) about the Sorting algorithm.

Edited by: Tree Sap on Jan 13, 2011 4:37 PM

[former_member192616]

> To fully answer my concern, I require confirmation (with a source,such as an article, book citation, book, etc.) about the Sorting algorithm.

are you interested in the algorithm (which i think is not published) or in the scalling behaviour (e.g. sort is n * log(n) like quick sort, binary search is log(n) linear search is n which is documented...)?

Kind regards,

Hermann

[Former Member]

Hi.

Indeed, I'm interested in the "scalling behaviour (e.g. sort is n * log(n) like quick sort, " /Big O/BigOmega performance behavior.

I'm required to make a mathematical analysis of the sorting performance used by SAP vesus certain amount of reads. I don't need to know the source code for the algorithm that SAP uses, but instead, the name of the algorithm or it's relative computational complexity(Big O, (e.g. sort is n * log(n) like quick sort)

e.g. "On XSAP book page 23, it is sugested that SAP uses something similar to quick sort" or "wikipedia.org/SAP_performance mentions that SAP has a computational complexity for sort in the order of 2n * log(n) +k"......

[former_member192616]

Hi,

> Indeed, I'm interested in the "scalling behaviour (e.g. sort is n * log(n) like quick sort, " /Big O/BigOmega performance behavior.

you can find them in standard sap courses (BC490, BC402) and in blogs by Siegfried and in books (e.g. http://www.sap-press.com/products/ABAP-Performance-Tuning.html or http://www.dpunkt.de/buecher/3096/performance-optimierung-von-abap%26reg%3B-programmen.html)

> I'm required to make a mathematical analysis of the sorting performance used by SAP vesus certain amount of reads.

As a rule of thumb if i remember right i learnt one should have ~50 reads before a sort pays off. Maybe Siegfried cann tell you more here.

Kind regards,

Hermann