Boost logo

Boost :

From: Kevin Lynch (krlynch_at_[hidden])
Date: 2006-07-10 16:04:24

Paul A Bristow wrote:
> | 3) This "normalizes" the interface for the calls to the distribution
> | functions - every call for "P" has exactly one argument, and
> | not two or three or four depending on the distribution in use.
> How would you envisage this working with Fisher, for example which has
> degrees of freedom 1 and 2, and a variance ratio.
> Is this a 1D or 2D or 3D?
> Its inversion will return df1 (given df2 and F and Probability)
> or df2 (given df1, F and Prob)
> or F (given Df1 and df2 and Prob)
> WOuld you like to flesh out how you suggest handling all these?

I can't say that I would :-) This is going somewhat beyond my area of
expertise ... it may well be the case that my model is just wrong, but
happens to work within the subset of probability and statistics that
I've had cause to use. I'd be perfectly willing to hear that criticism.

Some questions from the clueless to help me think through this:

Is it really the case that df1 in the Fisher distribution is any
different than, say, the mean or deviation in the Normal? For example,
I wouldn't talk about inverting the normal and extracting the mean. Or
would I? That's not something I've ever thought about. In my work, I
wouldn't try to extract the number of degrees of freedom for my
Chi-squared distribution, but then again, I only do fits with same, in
which case I always know the dof a priori. I guess I should really read
through that Probability and Statistics text I bought but haven't gotten
around to reading :-)

Kevin Lynch				voice:	(617) 353-6025
Physics Department			Fax: (617) 353-9393
Boston University			office:	 PRB-361
590 Commonwealth Ave.			e-mail:	 krlynch_at_[hidden]
Boston, MA 02215 USA

Boost list run by bdawes at, gregod at, cpdaniel at, john at