Curve Fitting

Scripting

ORCAScripts can be used to automate and use some of the analysis functions, either directly using the values from the simple statistics collected by the ORCA analysis tool or by using the more powerful fitting services provided by by OrcaROOT. If running a script against multiple plots, see the GOTCHA 

in the example below.

An ORCAScript example to do a quadratic fit on one of the DataGen plots:

function main()

{

//find the dataset

ds = find(ORHistoModel,1);

dataset = [ds dataSetWithName:"DataGen,0,1"];

//get the dataset dialog and open the analysis drawer

[dataset makeMainController];

plot = [dataset findController];

[plot openAnalysisDrawer];

//set up the region of interest

roi = [plot curve:0 roi:0];

[roi setMinChannel:100];

[roi setMaxChannel:550];

sleep(1); //wait a bit for the drawer to open and the analysis to be done

//find the fitter and do a arbitrary fit

fitter = [[plot fitController] model];

[fitter doArbitraryFit:"[0]+[1]*x+[2]*x^2"];

if(waituntil([fitter fitValid],5)){

n = [fitter fitParamCount];

for(i=0;i<n;i++){

print [fitter fitParamName:i],"=",[fitter fitParam:i],"+/-",[fitter fitParamError:i];

}

print "Chi Square: ",[fitter chiSquare];

}

    [plot close]; //POTENTIAL GOTCHA -- if you are working with more that one plot

                  //you must close the window after getting the analysis values

                  //otherwise if you have ORCA set to share windows then the wrong

                  //analysis drawer will be attached to the window.

                  //Closing the window resets everything

Running this script resulted in the following output in the status log:

010610 13:49:55 [OrcaScript] p0 = -32.0722 +/- 1.57874

010610 13:49:55 [OrcaScript] p1 = 0.274086 +/- 0.00934255

010610 13:49:55 [OrcaScript] p2 = -0.000382105 +/- 0.000012418

010610 13:49:55 [OrcaScript] Chi Square:  27.8595

Some knowledge of the types of fit parameters returned from ROOT is needed to determine which functions to call and which values to grab when the fit is done.

Some of the fitting functions you can call:

[fitter doGaussianFit];

[fitter doLandauFit];

[fitter doExponentialFit];

[fitter doPolynomialFit: fitOrder] -- does a polynomial fit of any order

[fitter doArbitraryFit:”ROOT-Type fit function”]  -- fit a function like “[0] + [1]*x + [2]*x^x”. Use the standard ROOT format for the function string.

[fitter doFitType: fitType] --where fitTypes are gaussian = 0, exponential = 1, landau = 3. Note that fitTypes 2 and 4 are covered by the doLinearFit and doArbitraryFit functions. The fitType is the index of the fit popup item in the analysis drawer. Use this if we add additional fits to the set and don’t include a separate fit method to call.

[fitter doFitType: fitType fitOrder:order fitFunction:”ROOT-Type fit function”] --same as above but giving more options. The fitType is the index of the fit popup item in the analysis drawer. Use this if we add additional fits to the set and don’t include a separate fit method to call.

Markers