Hi,
I have a group of process performance data and distribution grouped by condition. I am looking for an efficient way to calculate min, max, mean and std dev of the range of each condition group of data. Is there any platform in JMP that can do this?

Thanks,
Gary