Re: RFC: a proposed alternative solution to the "plotting columns from multiple files" problem

[<pl...@pi...>]

On 09/04/13 21:33, sfeam <Ethan A Merritt> wrote:
> The thing is, I don't think your proposal would actually work in practice
> for the "fit" part of this.  If every cycle of L-M fitting requres re-reading
> the original files through an intermediate layer of data presentation,
> my guess is that the throughput would become so awful as to be
> unusable.


It's true, iteratively doing plot and fit is going to be cumbersome 
whatever the details are.

Now, sorry if I'm mis-understanding the object of the task to be done 
here but isn't 'finding the best scaling' to match two files really 
regressing one ordinate series against the other one?

This raises the methodological issue of error/uncertainty in the mantissa.

Unless one is the controlled variable the resulting fit will be in 
error. (Eg. regressing a linear model will under-estimate the slope in 
the presence of x errors).

Assuming a fixed linear scaling is appropriate, doing the regression 
then inverting the data and fitting the other way around permits a first 
approximation (eg geometric average of the two slopes) but it really 
requires some added knowledge of the data. Non-trivial problem.

/Peter.