Using Wikidata to gather population data of cities and countries. Then calculate the entropy of drawing a uniformly random person labelled with corresponding city/country. Compare to how many bits would be necessary to uniquely specify an individual in the total population of those selected locations.

Data bias of being on Wikidata in the first place. Locations without population fields are assumed to be 0 population. The values outputed are only supposed to give a relative idea as compared to entropies of uniform distribution on just the location and uniform distribution over individuals counted in those population totals.

If data is available do the same with more such as

• Birth Year
• Occupation

Likely data not available for combinations but should output the upper bound gotten by calcuating with independent marginals.