RooAbsAnaConvPdf

 Destructor

 Declare a basis function for use in this physics model. The string expression 
 must be a valid RooFormulVar expression representing the basis function, referring
 to the convolution variable as '@0', and any additional parameters (supplied in
 'params' as '@1','@2' etc.

 The return value is a unique identifier code, that will be passed to coefficient()
 to identify the basis function for which the coefficient is requested. If the
 resolution model used does not support the declared basis function, code -1 is
 returned. 

 Change the resolution model to given model

 All direct generation of convolution arg if model is truth model

 Return a pointer to the convolution variable instance used in the resolution model

 Calculate the current unnormalized value of the PDF

 PDF = sum_k coef_k * [ basis_k (x) ResModel ]

 Handle trivial no-integration scenario

 Return analytical integral defined by given scenario code.

 For unnormalized integrals this is
                    _                _     
   PDF = sum_k Int(dx) coef_k * Int(dy) [ basis_k (x) ResModel ].
       _
 where x is the set of coefficient dependents to be integrated
 and y the set of basis function dependents to be integrated. 

 For normalized integrals this becomes

         sum_k Int(dx) coef_k * Int(dy) [ basis_k (x) ResModel ].
  PDF =  --------------------------------------------------------
         sum_k Int(dv) coef_k * Int(dw) [ basis_k (x) ResModel ].

 where x is the set of coefficient dependents to be integrated,
 y the set of basis function dependents to be integrated,
 v is the set of coefficient dependents over which is normalized and
 w is the set of basis function dependents over which is normalized.

 Set x must be contained in v and set y must be contained in w.

 Default implementation of function advertising integration capabilities: no integrals
 are advertised.

 Default implementation of function implementing advertised integrals. Only
 the pass-through scenario (no integration) is implemented.

 This function forces RooRealIntegral to offer all integration dependents
 to RooAbsAnaConvPdf::getAnalyticalIntegralWN() for consideration for
 analytical integration, if RRI considers this to be unsafe (e.g. due
 to hidden Jacobian terms). 

 RooAbsAnaConvPdf will not attempt to actually integrate all these dependents
 but feed them to the resolution models integration interface, which will
 make the final determination on how to integrate these dependents.

 Build complete list of coefficient variables 

 Throw away cache of coefCoefVars and normalization integrals constructed in terms of those coefficients

 Overload of hook function in RooAbsPdf::syncNormalization(). This functions serves
 two purposes: 

   - Modify default normalization behaviour of RooAbsPdf: integration requests over
     unrelated variables are properly executed (introducing a trivial multiplication
     for each unrelated dependent). This is necessary if composite resolution models
     are used in which the components do not necessarily all have the same set
     of dependents.

   - Built the sub set of normalization dependents that is contained the basis function
     resolution model convolution (to be used in syncNormalizationPostHook().

 Overload of hook function in RooAbsPdf::syncNormalization(). This function propagates
 the syncNormalization() call to all basis-function/resolution-model convolution component
 objects and fixes the physics models client-server links by adding each variable that
 serves any of the convolution objects normalizations. PDFs by default have all client-server
 links that control the unnormalized value (as returned by evaluate()), but convoluted PDFs
 have a non-trivial normalization term that may introduce dependencies on additional server
 that exclusively appear in the normalization.

 Print info about this object to the specified stream. In addition to the info
 from RooAbsPdf::printToStream() we add:

   Verbose : detailed information on convolution integrals

 Coefficient normalization access