Consider adding df_stats2 for 2d statistical functions like cor()

[rpruim]

The may require writing df_aggregate2() since aggregate() and hence df_aggregate() are intended for functions of one variable, not two.

Other functions this would help with include diffprop() and diffmean()