git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: Generalised bisection

From
STSteven Tweed <orthochronous@gmail.com>
Date
Mar 12, 2009, 18:02 UTC
Message-ID
<d9c1caea0903121102y5452603fua0e7a1b82e121b01@mail.gmail.com>
In-Reply-To
<alpine.DEB.1.00.0903121154560.10279@pacific.mpi-cbg.de>

On Thu, Mar 12, 2009 at 10:55 AM, Johannes Schindelin <Johannes.Schindelin@gmx.de> wrote:

Show 11 quoted lines
> On Thu, 12 Mar 2009, John Tapsell wrote:
> > 2009/3/11 Ealdwulf Wuffinga <ealdwulf@googlemail.com>:
> > > What I use is the multiprecision floating point number class. doubles
> > > don't seem to be long enough.
> >
> > Hmm, really really?  Sometimes this sort of thing can be fixed by just
> > readjusting the formulas.  What formulas are you using that require more
> > precision than doubles?
>
> Maybe you could post the formulae instead of forcing people to deduct them
> from the source code?

I haven't even looked at the source code so a description of the mathematical algorithm would help, but I'll just point out that underflow (in the case of working with probabilities) and overflow (when working with their negated logarithms) is inherent in most multi-step Bayesian algorithms. The only solution is to rescale things as you go so that things stay in a "computable" range. (You're almost never interested in absolute probabilities anyway but rather relative probabilities or, in extreme cases, just the biggest probability, so rescaling isn't losing any useful information.)

cheers, dave tweed

Previous: Johannes SchindelinNext: Ealdwulf Wuffinga
Message 13 of 25 in “Generalised bisection”
  1. Ealdwulf WuffingaMar 9, 2009
  2. Christian CouderMar 10, 2009
  3. Ealdwulf WuffingaMar 11, 2009
  4. John TapsellMar 11, 2009
  5. Johannes SchindelinMar 11, 2009
  6. John TapsellMar 11, 2009
  7. Johannes SchindelinMar 11, 2009
  8. John TapsellMar 11, 2009
  9. Ealdwulf WuffingaMar 11, 2009
  10. Ealdwulf WuffingaMar 11, 2009
  11. John TapsellMar 12, 2009
  12. Johannes SchindelinMar 12, 2009
  13. Steven TweedMar 12, 2009
  14. Ealdwulf WuffingaMar 13, 2009
  15. Ealdwulf WuffingaMar 13, 2009
  16. Steven TweedMar 13, 2009
  17. Ealdwulf WuffingaMar 15, 2009
  18. Steven TweedMar 16, 2009
  19. John TapsellMar 16, 2009
  20. Ealdwulf WuffingaMar 16, 2009
  21. Ealdwulf WuffingaMar 16, 2009
  22. Ealdwulf WuffingaMar 13, 2009
  23. Johannes SchindelinMar 13, 2009
  24. John TapsellMar 13, 2009
  25. Johannes SchindelinMar 13, 2009

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.