Comment on Information Gain and Mutual Information for Machine Learning by Jason Brownlee

You’re welcome.
S and a are two “variables” just like X and Y. We perform the same calculation on a split of the data as we do on two columns in a dataset.