Re: Implementation of Grid File

Hi, I have a grid that is 4 columns wide and n row's deep. I need it to be very efficient, as they're could be millions of rows, so secondary storage efficiency is needed.

      A
     B
     C
     D

      101
     1
     5
     2

      102
     101
     6
     3

      103
     102
     7
     4

      104
     1
     8
     4

We may have 1 or 2 value's to query for a row, for example we could have D & C and would need to find the row that matches, and more than 1 row could match the criteria.To my knowledge, a grid file data structure provides this function.

Regards

John

"Benjamin Johnston" <superhero_at_benjaminjohnston.com.au> wrote in message news:BkVQ9.14961$jM5.41661_at_newsfeeds.bigpond.com...
> > Hi,
> >
> > Does anyone know of any source code for a grid file, preferably java?
I've
> > search the internet and newsgroups, but they're doesn't seem to much
> around.
>
> Unfortunately, it looks like the grid file is a bit too obscure a database
> data structure to be freely available AND well known (I couldn't find any
> Java implementation).
>
> I expect it would not be too hard to implement yourself. Just keep the
axes
> in memory, and use a random access file to read the index.
>
> Alternately, depending on your application, it might be possible to use a
> Quadtree (type "java quadtree" into google without the quotes, for many
free
> implementations). If you're storing points you could use a Z-ordering and
> stick your data into a standard database. A quadtree or any other kind of
> index might not be perfectly optimal for your particular application, but
it
> could be "good enough" for your purposes.
>
> So why did you select a grid file, and exactly what kind of data were you
> intending to store in the grid file?
>
> -Benjamin Johnston
>
>
>
Received on Fri Jan 03 2003 - 00:22:44 CET