I'm against this. Either you have to create a new statistics module or you would have to include numpy/pandas/statsmodels into the standard library. In both cas

I'm against this. Either you have to create a new statistics module or you would have to include numpy/pandas/statsmodels into the standard library. In both cases it would essentially freeze the modules for further development outside the python release cycle...