Refactor age-related variables

[MaxGhenis]

variables/demographic/person.py includes four age-related variables:

• over_16: person("age", period) >= 16
• is_adult: person("age", period) >= 18
• is_child: person("age", period) < 18
• is_WA_adult: person("is_adult", period) * not_(person("is_SP_age", period))
• is_young_child: person("age", period.this_year) < 14
• age_under_18: person("age", period) < 18
• age_18_64: (age >= 18) & (age <= 64)
• age_over_64: person("age", period) > 64
• is_older_child: (person("age", period) >= 14) * (person("age", period) < 18)

One issue is that over_16 is inconsistently defined with over_64 (>= vs >). I think a couple of these can also use between().

But more generally, terms like young_child and even child can be ambiguous (sometimes a person under 18 isn't a child if they're not in a family). Can we remove or rename the variables that don't specify their criteria? We could also set a standard around only defining under_* and between_*_* to avoid redundancy. For example, all the above could be shrunk down to:

• under_14
• between_15_17
• under_17
• under_18
• between_18_SPA
• between_18_64
• under_65

[MaxGhenis]

@nikhilwoodruff wdyt here? This is mostly unimportant but I think over_16 = person("age", period) >= 16 is sort of a naming bug.

[MaxGhenis]

Renaming and shifting gears on this. We should instead align the age-level classifications to the law, and avoid parameter values in these variables.