Thanks for this package. I noticed however, when I was checking the Goodman-Kruskal Lambda and some others the results were slightly different as when I would calculate them 'by hand'. I did some digging and it seems you add a +1 to each cell count in the contingency table. I think the line of code doing this is:

Creating the table

table = [[df.query(f'a=="{x}" and b=="{y}"').shape[0] + 1 for y in b_values] for x in a_values]

Is there a reason why this +1 always gets added, and should it?