Add experimental design functions such as market selection for control and treatment

[Starbz]

Hi,

As described in the medium article's commentary, would be super helpful if we could add some experimental design functions to compute the best treatment & control groups to then build the best synthetic control possible. Furthermore building this kind of power analysis could also help assess a minimum test duration, a minimum detectable effect etc.

Just an idea

Arthur