On Sunday 15 March 2009, you wrote:
Dear Arno,
> I must admit that I also find it hard to understand what this function
> is useful for. Wouldn't you rather want a function that either 1)
> computes confidence intervals for given confidence levels; or 2) tests
> if samples come from the multinomial distribution (similar to a
> Lilliefors test for the normal distribution).
Almost. But not exactly. At Lilliefors you want to decide if it comes from a
specific distribution
namely Gaussian distr.
Here I do calculate the level of confidence at which the samples represent the
true distribution
given that there is a tolerance (confidence interval).
This the upside down case of the typical excercises at which we want to get the
confidence interval
given the confidence level (and the estimated parameters of the distribution).
But once we accept (lets say at elections) that we have a standard predefined
maximal acceptable error rate in the estimation and we just want to know that
how sure we
can be that the measured proportions are the same as in the
entire population (ie. the expected value and mean of the samples are the same)
we need to do this.
>
> It would help to see an example for which the function provides a
> meaningful solution. The references do not provide such an example as
> their ultimate goal is to determine minimum samples sizes / to validate
> approximate confidence intervals.
The references explain the relationship between samples (N) (from multinomial),
confid interval (CI) and confid levels (CL).
Sisan and Graz presented a much more accurate
solution but since its non-compact solution it was impossible to reformulate
it and express the CL from the other two, I turned to older results.
>
> General suggestions for improvements:
>
> - As Soren pointed out, the help text is hard to understand. Apart of
> the issue with the uncommon objective, it is confusing to use term
> "cell" which has another meaning in Octave programming. I think
> "category" would be better.
Ok.
>
> - You briefly mention that the "agresti_cull" option for
> calculation_type is a modification of the method that is described in
> the corresponding reference. It would be helpful to elaborate this a bit
> more so that it is clear what the option does.
Well, these guys make first an improvement which is mostly based on their
gut feelings than test. Not much explanations.
But recently I got the article (I had only a review from this) will go through
if
find something reasonable.
> - Argument checks help avoiding bugs
Yes. But maybe I should use opt package instead.
This is what we recently talked with Jaroslav.
Lev
--
Blogger of http://fapuma.blogspot.com
------------------------------------------------------------------------------
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
easily build your RIAs with Flex Builder, the Eclipse(TM)based development
software that enables intelligent coding and step-through debugging.
Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
_______________________________________________
Octave-dev mailing list
Octave-dev@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/octave-dev
