Re: yet another hierarchy model

kwillets_at_looksmart.net (KWillets) wrote in message news:<eef24980.0110161354.7eb13a84_at_posting.google.com>...
> Vadim Tropashko <nospam_at_newsranger.com> wrote in message news:<Urmx7.24985$ev2.33971_at_www.newsranger.com>...
> > Here is one more way to represent hierarchy: just store depth-first sequence
> > number and level (simple, indeed!).
> >
> (...until one needs to delete or insert a node.)
>
> How about:
>
> 1. Assigning unique prime numbers to all leaf nodes, and assigning
> each intermediate node the product of its subordinates.
>
<snip>
>
> This kind of thing can go on and on....

Indeed. Here are 2 more methods:

1. Modified Celko's nested sets:

.      (1,1)
.      / | \  
.     /  |  \  
.    /   |   \  
. (1,3)(2,2)(3,1)
.              \
.               \
.              (4,1)

(i,j) is a child of (k,l) iff i>=k and j>=l

2. Inverted Willets' division lattice

.        1
.       /|\ 
.      / | \  
.     /  |  \  
.    /   |   \  
.   2    3    5
.  /|\  /|\  
. 4 6 8/ | \ 
.     /  |  \ 
.    9  15  21 

In multiples of the node '3' we omit multiples of the node '2'.

Looks like any partial order would do. Received on Thu Oct 25 2001 - 09:04:26 CEST