Add Group Analytics data to seed script

[simfish85]

We could add 'Family' as a group type, letting multiple members be part of the same family. To keep things less random we could have a fixed list of (say) 200 family names and possibly 1000 users split across those families randomly which are stored as a CSV and then they are randomly selected when generating the historic data.

We use Faker live to generate the data but think we could also use it to create a CSV to keep the names and groups fixed https://faker.readthedocs.io/en/stable/providers/faker.providers.person.html#faker.providers.person.Provider.last_name