Thread: method to extrapolate groups in a given data set

Dear all:
Is there a software that can mine a dataset and extrapolate and organize its data into groups.
For example. Lets say we know that Athletes drinks 2x more water than couch potatoes.

A study was conducted blind and a set of data was given as such.

subject 1: 1 liter
subject 2: 1.2 liter
subject 3: 500 ml
subject 4: 2 liter
subject 5: 200 ml
subject 6: 700 ml

A computer mining program will then group
subjects 1,2,4 as group A
subjects 3, 5, 6 as group B

Of course in the real world we do not necessarily know that two groups exist in this given data set, or even that athletes drink more water than couch potatoes.

thanks!
Alex :wink:

The easy way would be to sort the data by the amount of water consumed. Then you would have to specify some criteria as a breakpoint between the groups. It is not clear to me why you grouped the six people as you did, unless you just arbitrarily decided to put the cutoff at 1 liter. The two groups are not athletes and couch potatoes as you have defined them.