Skip main navigation (Press Enter).
Log in
Toggle navigation
Log in
Community
Topic Groups
Champions
Directory
Program overview
Rising Champions
IBM Champions group
User Groups
Directory
Benefits
Events
Dev Days
Conference
Community events
User Groups events
All TechXchange events
Participate
TechXchange Group
Welcome Corner
Blogging
Member directory
Community leaders
Resources
IBM TechXchange
Community
Conference
Events
IBM Developer
IBM Training
IBM TechXchange
Community
Conference
Events
IBM Developer
IBM Training
Global AI and Data Science
×
Global AI & Data Science
Train, tune and distribute models with generative AI and machine learning capabilities
Group Home
Threads
4K
Blogs
911
Events
0
Library
370
Members
28.5K
View Only
Share
Share on LinkedIn
Share on X
Share on Facebook
Back to Blog List
Generating Random Latin Squares
By
Moloy De
posted
Thu November 05, 2020 09:24 PM
Like
In one of my previous blogs
here
I talked about how the area of Latin Square evolved from Statistical Design of Experiments. Latin Squares due to its simple yet rich structures have fascinated mathematicians like Euler and others. Construction of a pair 10 by 10 Mutually Orthogonal Latin Squares have been a milestone work by Prof. Bose, Prof. Srikhande and Prof. Parker in 1959. In this blog I talk about a general method of constructing a Latin Square of order n.
A Latin Square is an n by n arrangement of n symbols, say the numbers 0 to n-1, where there is no duplicate in any row and in any column. Below is an example of 8 by 8 Latin Square:
Since each row is an arrangement of n distinct objects / numbers, each row is a permutation and could be mapped to an element of a permutation group S
n
. Those who are interested to study the details about the Permutation Group are referred to the Wikipedia Page
here
. If we consider an acyclic permutation a, then {1, a, a
2
, ..., a
n-1
} form a subgroup of S
n
and they form the rows of an n by n Latin Square. Clearly no row has any duplicate since they are arrangements of distinct numbers. No column has a duplicate either. As otherwise the starting permutation a fails to be acyclic.
Clearly these do not include all the Latin Squares of Order n. In fact rows of an order n Latin Square need not even be a group. Still we may relax the condition of the starting permutation a to be acyclic and may consider to be product of r cycles each of length s where n = rs. Then also it is possible to work out n permutations that form a subgroup of S
n
and eventually form the rows of a Latin Square of order n.
So if we start we a random permutation of n numbers, create the corresponding natural acyclic permutation out of it, then we get a random Latin Square of order n by considering its powers. Till now I don't know any other method of generating random Latin Squares more efficient than this.
Extendibility of such Latin Squares to its Orthogonal Mate is a different question.
QUESTION I: How is the possibility of a Latin Rectangle?
QUESTION II: How many Latin Squares of order n are there?
#GlobalAIandDataScience
#GlobalDataScience
0 comments
2 views
Permalink
Copy
https://community.ibm.com/community/user/blogs/moloy-de1/2020/11/05/points-to-ponder
Powered by Higher Logic