Interesting comparisons can be made, purely using SQL, by calculating the coefficient of variation for all of the variables, and grouping by outcome. A coefficient of variation is a way in which we can standardize the means and variances of the variables and thus be able to disregard scale when making comparisons:
%sql select outcome, mean(pregnant)/std(pregnant), mean(glucose)/std(glucose), mean(pressure)/std(pressure), mean(triceps)/std(triceps), mean(insulin)/std(insulin), mean(mass)/std(mass), mean(pedigree)/std(pedigree), mean(age)/std(age) from out_tbl group by 1
After the query is done, switch to the plot window, and change the plot type to line chart.
Then drag the outcome ...