class RooXYChi2Var: public RooAbsOptTestStatistic

  RooXYChi2Var constructor with function and X-Y values dataset

 An X-Y dataset is a weighted dataset with one or more observables X where the weight is interpreted
 as the Y value and the weight error is interpreted as the Y value error. The weight must have an
 non-zero error defined at each point for the chi^2 calculation to be meaningful.

 To store errors associated with the x and y values in a RooDataSet, call RooRealVar::setAttribute("StoreError")
 on each X-type observable for which the error should be stored and add datapoints to the dataset as follows

 RooDataSet::add(xset,yval,yerr) where xset is the RooArgSet of x observables (with or without errors) and yval and yerr
                                 are the Double_t values that correspond to the Y and its error

  RooXYChi2Var constructor with function and X-Y values dataset

 An X-Y dataset is a weighted dataset with one or more observables X where given yvar is interpreted
 as the Y value. The Y variable must have a non-zero error defined at each point for the chi^2 calculation to be meaningful.

 To store errors associated with the x and y values in a RooDataSet, call RooRealVar::setAttribute("StoreError")
 on each X-type observable for which the error should be stored and add datapoints to the dataset as follows

 RooDataSet::add(xset,yval,yerr) where xset is the RooArgSet of x observables (with or without errors) and yval and yerr
                                 are the Double_t values that correspond to the Y and its error

 RooXYChi2Var constructor with an extended p.d.f. and X-Y values dataset
 The value of the function that defines the chi^2 in this form is takes as
 the p.d.f. times the expected number of events

 An X-Y dataset is a weighted dataset with one or more observables X where the weight is interpreted
 as the Y value and the weight error is interpreted as the Y value error. The weight must have an
 non-zero error defined at each point for the chi^2 calculation to be meaningful.

 To store errors associated with the x and y values in a RooDataSet, call RooRealVar::setAttribute("StoreError")
 on each X-type observable for which the error should be stored and add datapoints to the dataset as follows

 RooDataSet::add(xset,yval,yerr) where xset is the RooArgSet of x observables (with or without errors) and yval and yerr
                                 are the Double_t values that correspond to the Y and its error

 RooXYChi2Var constructor with an extended p.d.f. and X-Y values dataset
 The value of the function that defines the chi^2 in this form is takes as
 the p.d.f. times the expected number of events

 An X-Y dataset is a weighted dataset with one or more observables X where the weight is interpreted
 as the Y value and the weight error is interpreted as the Y value error. The weight must have an
 non-zero error defined at each point for the chi^2 calculation to be meaningful.

 To store errors associated with the x and y values in a RooDataSet, call RooRealVar::setAttribute("StoreError")
 on each X-type observable for which the error should be stored and add datapoints to the dataset as follows

 RooDataSet::add(xset,yval,yerr) where xset is the RooArgSet of x observables (with or without errors) and yval and yerr
                                 are the Double_t values that correspond to the Y and its error

 Copy constructor

 Common constructor initialization

 Initialize bin content integrator

 Destructor

 Calculate contribution to internal error due to error on 'x' coordinates
 at point i

 Return function value requested bu present configuration

 If integration is required, the function value integrated
 over the bin volume divided by the bin volume is returned,
 otherwise the value at the bin center is returned.
 The bin volume is defined by the error on the 'X' coordinates

 If an extended p.d.f. is used as function, its value is
 also multiplied by the expected number of events here

 Calculate chi^2 in partition from firstEvent to lastEvent using given stepSize

 Inform base class that observable yvar cannot be optimized away from the dataset

 Virtual constructor

 The default error level for MINUIT error analysis for a chi^2 is 1.0

 Disable function (component) caching if integration is requested as the function
 will be evaluated at coordinates other than the points in the dataset