some improvements

[choonghyunryu]

from @roberto Passera ,

1- descriptive univariate analysis for continuous variables, it might be useful to add standard deviation too (a classical evergreen measure…)
2- descriptive univariate analysis for categorical variables, I do believe that a classical vertical bar plot could be more immediate, letting it to be copied and pasted as the other graphics do
3- bivariate categorical variables (and everywhere), what about an option to exclude missing from the general count?
4- Target based Analysis - Grouped Categorical Variables, the percentages for relative contingency tables do not go, at least with my datasets