Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: nonlinear-leastsquare


From   "Nick Cox" <[email protected]>
To   <[email protected]>
Subject   st: RE: nonlinear-leastsquare
Date   Mon, 1 Dec 2003 11:24:25 -0000

Andreas Aschbacher
 
> I have two columns in ascii-sheet :
> x-col        y-col
> 1              34
> 2              45
> 3              44
> ... and so on         /experiment in radioactive measurement /
> 
> and the following program in a do-file:
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> capture program drop nlexample
> program nlexample
> version 8
> if "`1'" == "?" {
> global S_1 "a1 a2 a3 a4 a5 a6"

... 

> I reached convergence and good values for a1,a2,...with 
> "right"start-values
> above after long 'trying' ......

> my questions::
> 1.) which parameter is the typical Marquardt-Parameter ?

Pass. 

> 2.) although in reference-book:what does query-call and what does
> calculation-call mean ?

Think of the query call as initialising the problem and 
the calculation calls as doing the numerics and statistics. 

> 3.)after calling the program above with >do name< and after 
> getting my
> results I print: predict myvar
>     then I  have  with the command :twoway line myvar x a passable
> graph-fit,is this the correct way ?

This plots predicted against predictor. 

Given that curve, 

. twoway mspline myvar x 

might look better. 

You might also want to add the data points, say by 

. twoway mspline myvar x || scatter y x 

Alternatively, 

. ssc desc regplot 
. ssc inst regplot 
. <nl fit> 
. regplot y 

Nick 
[email protected] 

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index