Capability analysis CPK vs PPK calculation?

The S = capability(data,specs) function provides a capability analysis with resultant CPK (capability within subgroups), is there a way to calculate PPK (overall capability) for a data set?

Answer 1

[Abhaya]

Hi,

I understand you want to calculate process performance index.

I have calculated the Ppk value for vector ‘x’ using following code. I have set the lower specification limit (LSL) = 9.8 and upper specification limit (USL) = 10.2 .

The code is as follows:

function S=ppkCalc(data,spec) 
    %claculate mean 
    S.mu=mean(mean(data)); 
 
    %find the sum of square of each number 
    sumSq=sumsqr(data); 
 
    %find standard deviation 
    sz=size(data); 
    n=sz(1)*sz(2); 
    S.sigma=sqrt((sumSq-n*(S.mu^2))/(n-1)); 
 
    LSL=spec(1); 
    USL=spec(2); 
 
    S.Ppk=min((S.mu-LSL)/(3*S.sigma),(USL-S.mu)/(3*S.sigma)); 
end
% the following measurement are for five different objects with target of 10 and tolerance 0.2 
mat=[10.1	10.0	9.9	10.2	10.0;
10.0	10.1	9.8	10.2	10.1; 
10.3	10.0	10.1	9.9	10.2; 
10.0	9.9	10.1	10.0	10.2; 
9.8	10.1	10.0	10.2	10.0]; 

%call the function
x=ppkCalc(mat,[9.8 10.2]) 

x = struct with fields:
       mu: 10.0480
    sigma: 0.1295
      Ppk: 0.3913