Hello,

Sorry if this has been asked before, it is probably a common problem but I haven’t been able to find it in search.

I am making a fit with 4 parameters and want to make a plot where I have parameter 0 in x and parameter 1 in y.

Since they are correlated, I need something like a TGraphBentErrors, I think (maybe there is a more automatic way of doing this?). The problem is that I haven’t found any documentation and don’t know the meaning of the last 4 parameters. After trial and error I found that

```
protao1900[4]->Fit("ghRL","QNI","",-300,200);
r_r0[0]=ghRL->GetParameter(0);
l_r0[0]=ghRL->GetParameter(1);
er_r0[0]=ghRL->GetParError(0);
el_r0[0]=ghRL->GetParError(1);
Double_t matrixp[4][4];
gMinuit->mnemat(&matrixp[0][0],4);
tiltxl[0]=-matrixp[0][1]/sqrt(matrixp[0][0]*matrixp[1][1]);
tiltxh[0]=matrixp[0][1]/sqrt(matrixp[0][0]*matrixp[1][1]);
....
TGraphBentErrors *r_0=new TGraphBentErrors(3,r_r0,l_r0,er_r0,er_r0,el_r0,el_r0,tiltxl,tiltxh,0,0);
```

gives approximately what I want as far as the angle of the x errors bars with respect to the horizontal is concerned, however I don’t know what to put in the y errors.

Also, could you expand a bit on the TGraphBentErrors page and section in the guide? It just gives two values of 0.05 for the last parameters and shows two error bars making approximately 45º with the other errors. Maybe it is somewhere and it is very obvious, but I couldn’t understand it.

Thank you