diff --git a/CCREClass.m b/CCREClass.m
new file mode 100644
index 0000000..550694c
--- /dev/null
+++ b/CCREClass.m
@@ -0,0 +1,77 @@
+classdef CCREClass<CREClass 
+    %CCRECLASS calculate the Conditional Cummulative Residual Entropy
+    %   This class uses the CREClass options
+    %   
+    
+    properties
+        %the data field inherited from the CREClass will contain the data
+        %to calculate the CCRE on. 
+        %the nBin field inherited from the CREClass will contain the number
+        %of bins if UseHistProxy is set. 
+    end
+    properties
+        DataRef %This parameter will contain the reference set.
+        nBinRef
+        CCRE %parameter containing the conditional Cumulative Residual Entropy
+    end
+    
+    methods %constructor
+        function obj = CCREClass()
+            %CCRECLASS Construct an instance of this class
+            %   Detailed explanation goes here
+            obj@CREClass;
+        end
+    end
+    methods %set and get functions
+        function set.DataRef(obj,val)
+            if isnumeric(val)
+                obj.DataRef=val;
+            else
+                error('please enter numeric array/matrix as reference');
+            end
+        end
+        function set.nBinRef(obj,val)
+            nBinFromCRE=obj.nBin; %park the current value of nBin
+            obj.nBin=val; %use the check system from CREClass
+            obj.nBinRef=obj.nBin;
+            obj.nBin=nBinFromCRE; %restore the current value of nBin
+        end
+    end
+    methods %user callable functions
+        %% CCRE A|B
+        %
+        % $$CCRE = \varepsilon (A) - E[\varepsilon(A|B)]$$
+        %
+        % $$\varepsilon(A) = -\sum_{\lambda}F_c^A(\lambda)logF_c(A^(\lambda)$$
+        % $$F_c^A(\lambda)=\int_{\lambda}^{\infty}p^A(l)dl$$
+        %
+        % $$E[\varepsilon(A|B)]= \sigma_{\kappa} p^B(\kappa)*\varepsilon(A|B)$$
+        %
+        % $$\varepsilon(A|B) = \sigma_{\lambda}F_c^{A|B}(\lambda)log
+        % F_c^{A|B}(\lambda)$$
+        %
+        % $$F_c^{A|B}(\lambda) = \int_{\lambda}^{\infty}p^{A,B}(l,k)/p^B(\kappa)dl$$
+        %
+        % see equation 7 in wan and Vemuri
+        function Calc(obj)
+            Calc@CREClass(obj); %calculate the CRE for the data. 
+            if isempty(obj.DataRef);return;end %if no refdata given, return;
+            %calculate histogram for RefData
+            EVarEpsDataGivenRef=0; %clear previous calculations
+            [pDataRef,E]=histcounts(obj.DataRef(:),obj.nBinRef,'Normalization','pdf');
+            [JointHist,E1,E2]=histcounts2(obj.Data(:),obj.DataRef(:),[obj.nBin,obj.nBinRef],'Normalization','probability');
+            for k=1:numel(pDataRef) %loop over the bins in pDataRef NB: must be equal to the bins in JointHist for the Ref
+                if pDataRef(k)==0;continue;end% if no data at this line go to the next
+                ind=E(k)<=obj.DataRef(:)&obj.DataRef(:)<E(k+1);
+                CRE_DataGivenDataRef=CREClass;
+                CRE_DataGivenDataRef.Data=obj.Data(ind);
+                CRE_DataGivenDataRef.nBin=E1;
+                CRE_DataGivenDataRef.Calc;
+                EVarEpsDataGivenRef=EVarEpsDataGivenRef+pDataRef(k)*CRE_DataGivenDataRef.CRE; % I really have to check this line on merit. 
+            end
+            obj.CCRE=obj.CRE-EVarEpsDataGivenRef;
+
+        end
+    end
+end
+