Not all histograms are normalised.
For example, individuals by hair color:
- Blond 2 (Alex, Beckham)
- Gray 3 (Foo, Bar, Baz)
- ...
This histogram is not normalised, because individuals in result set are relevant.
Compare with repairs by car make:
- Fiat 81%
- Daewoo 18%
- Toyota 1%
This histogram is normalised to 100% because individual number of repairs is irrelevant.