
#1
Dec1912, 01:37 AM

P: 9

Hey all, I have been playing around with a special type of matrix and am wondering if anyone knows of some literature about it. I have been calling it a pseudoorthogonal matrix but would like to learn if it has a real name or if we can come up with a better name. The characteristics of the matrix are as follows:
1) The matrix is composed of only ones and zeros 2) Each row and each column have the same number of ones in it. (If there are 3 ones in each row/column then I call a 3rd order matrix) 3) Between any two rows, there is one and only one common column with a one. Here is an example of what I call a 3rd order pseudoorthogonal matrix. Let's call him 'M' 1 1 1 0 0 0 0 1 0 0 1 1 0 0 1 0 0 0 0 1 1 0 1 0 1 0 1 0 0 1 0 0 1 0 1 0 0 1 1 0 0 1 0 0 1 0 1 1 0 I call it a pseudo inverse because inv(M) = M/21/6 , i.e. with adding and multiplying by constants I can arrive at the inverse of M. Has anyone played with something like this before? I am hoping to gleen information to help me generate higher order matrices of this type. 



#2
Dec1912, 02:42 AM

P: 4,570

Hey orthogonal and welcome to the forums.
I've never played with this kind of thing (I haven't had to): do you have a reason for doing this: (pure curiosity or do you have an application in mind)? I don't know whether this would help but error correcting code matrices (in binary XOR) might share common properties with this (it's just a hunch and its probably wrong anyway, but you never know!). 



#3
Dec1912, 10:15 AM

P: 9

I am working on the applications but I found the matrix by analyzing the matching card game Spot it! It is a card game which has 8 symbols per card with one and only one matching symbol between any two cards. If you follow the link above you can play a demo.




#4
Dec1912, 05:57 PM

P: 4,570

Pseudoorthogonal matrix?
How about considering the eigenvalues to create a higher dimensional matrix (as well as the eigenvectors)?
http://en.wikipedia.org/wiki/Matrix_diagonalization http://en.wikipedia.org/wiki/Eigenva...her_properties 



#5
Dec3012, 11:21 PM

P: 9

I have some findings to report:
I have written a program which can generate up to order 6 successfully, but when my code attempts to do order 7 it chokes (24 hours + with no solution returned!). After googling around some more I found a poster presentation which describes the problem using mutually orthogonal Latin squares. It looks like I have some reading to do to catch up on all this higher order geometry stuff. 


Register to reply 
Related Discussions  
pseudo inverse matrix  Linear & Abstract Algebra  1  
Proving transpose of orthogonal matrix orthogonal  Calculus & Beyond Homework  4  
Orthogonal matrices prove: T is orthogonal iff [T]_bb is an orthogonal matrix  Calculus & Beyond Homework  0  
[SOLVED] Pseudo orthogonal group  General Physics  204 