[seqfan] Re: Seq triangles that are unimodular

[israel at math.ubc.ca]

Any upper or lower triangular matrix with +-1's on the diagonal has 
determinant +-1, so if its entries are integers so are its inverse's.

Cheers,
Robert

On Jul 14 2017, Peter Lawrence wrote:

> Pascal's triangle as a matrix turns out to be unimodular (inverse has all 
> integer entries) Stirling numbers of first kind as matrix and of second 
> kind are inverses of each other Fibonomal triangle as matrix is 
> unimodular, A010048, and A103910 Gaussian Binomial Triangles turn out to 
> be unimodular, e.g. A022166 and A135950
>
>Is there a theory of triangular unimodular matricies ?
>
> For a given rank they form a group under multiplication, is there a 
> finite set of generators (it would seem not) ?, or a simple infinite set 
> ?
>
>Any info or references would be appreciated
>
>Thanks,
>Peter A. Lawrence.
>
>--
>Seqfan Mailing list - http://list.seqfan.eu/
>
>