Re: question about index cost estimates

Tom Lane wrote:
>
> Jeff Hoffmann <jeff(at)propertykey(dot)com> writes:
> > does anybody on the list know anything about fractals & wavelets?
>
> Now there's an interesting idea: regard the stats as a lossy compression
> of the probability density of the original dataset. Hmm ... this
> doesn't do anything for the problem of computing the pdf cheaply to
> begin with, but it might help with storing it compactly in pg_statistic.
>
> regards, tom lane

yeah, that's exactly what i meant. you can probably tell math &
statistics aren't my strongest points. i'm trying to learn a little
about fractals because about all i knew before today is those pretty
little pictures. just doing a quick search for their use with
databases, though, i found at least one paper on selectivity estimates &
fractals which i'm going to read. just for your reference, the paper is
from VLDB and it called "Estimating the Selectivity of Spatial Queries
Using the 'Correlation' fractal Dimension". here's the link:
http://www.vldb.org/conf/1995/P299.PDF. i've only read the abstract,
but it sounds pretty promising. i'm starting to wish that i knew what
was going on -- this is getting to be interesting to me. i've been
using postgresql for years, but up until the last few days, i was trying
to treat it as a black box as much as possible -- i probably should have
tried getting involved a lot earlier...

jeff