Re: Fit x,y [2]

Valery Fine (fine@mail.cern.ch)
Sun, 10 Aug 1997 15:27:53 +0100


On 10 Aug 97 at 13:27, Voropaev S.G. wrote:

> Hi,
> I investigate my problem. New example:
>
> float x[20], y[20];
> for(int i=0; i<20; i++) {x[i]=(i+1)/21.0; y[i]=0.5*x[i]+0.2;};
> c1 = new TCanvas("c1","proba");
> TGraph *tg = new TGraph(20,x,y);
> tg->Draw("AP");
> tg->Fit("pol1"); // All is OK. I see all my point and long line.
>
> for(int i=0; i<20; i++) {x[i]=gRandom->Rndm(1); y[i]=0.5*x[i]+0.2;};
> delete tg; TGraph *tg = new TGraph(20,x,y); tg->Draw("AP");
> tg->Fit("pol1");
>
> Now I see all my point and very short line. The X range of line is
> 0.1-0.2!

But did you see you the graphic view properly ?
You may say "yes" but I'll insist "no". That's because of the
shape of your function special function: 0.5*x + 0.2, Change your
function to something else and draw it and WITH APL options to
see the REAL its shape. (For example 0.5*x**2 + .2). No ask some
human (not computer) to fit that "by hand".

>
> Question: Can I fit non-sorted x,y data point?
>

No one has to sort it first. Otherwise I have no idea how to fit
the data supplied even myself (and even it is the straight line).
Otherwise one needs the third array to define the order of his
points explicitly just to allow TMinuit class to sort his data.

Wiht my regards,
Valery
=================================================================
Dr. Valery Fine Telex : 911621 dubna su
-----------
LCTA/Joint Inst.for NuclearRes Phone : +7 09621 6 40 80
141980 Dubna, Moscow region Fax : +7 09621 6 51 45
Russia mailto:fine@main1.jinr.dubna.su

Dr. Valeri Faine
------------ Phone: +41 22 767 6468
CERN FAX : +41 22 767 7910
CH-1211 Geneva, 23 mailto:fine@mail.cern.ch
Switzerland http://nicewww.cern.ch/~fine