commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Gilles Sadowski <gil...@harfang.homelinux.org>
Subject Re: [Math] Question about "AbstractLeastSquaresOptimizer" code
Date Wed, 25 Apr 2012 13:41:24 GMT
Hi.

> looking at various references, it seems that the definition of the error as
> errors[i] = FastMath.sqrt(covar[i][i]) * c;
> or
> errors[i] = FastMath.sqrt(covar[i][i]);
> is a matter of convention.
> 
> I'm no statistician, but I gathered that the first expression is what
> is called Asymptotic Standard Error (see equations 34 and 35,
> http://mathworld.wolfram.com/LeastSquaresFitting.html). Meanwhile, NR
> seems to favor the second expression. It would be nice to check what R
> (which is truly statistician's stuff) returns as parameters errors.

Thanks for looking into that.

Can we then agree that the current naming is not really clear?
Since we are not allowed to readily remove this method, the best solution
would be to update the Javadoc to make it clear what is done there, i.e.
cite the full definition of ASE (together with a reference).

Then, to further disambiguate, we could add a
  double[] getSigma()
method that will in effect return the (sqrt of the) diagonal elements of the
covariance matrix.

Is that OK?
If so, would you take care of that, S├ębastien?


Thanks,
Gilles

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org


Mime
View raw message