Is there any way in JMP where I can group data and get an output of goodness of fit and cpk (instead of going on by one), I just need to do a lot by lot analysis only for goodness of fit and cpk on more than 1000 lots
Thanks for your help