470 integer child ages

[paddy-r]

Do...

• baseline, general check -> runs without errors
• handovers, see that child ages and nkids (hh) look okay, maybe compare with development handovers, should be identical -> see images below, some differences -> not identical but very similar
• Do spot checks on child ages -> marking as done because not necessary as child age plots very similar to development

[paddy-r]

@RobertClay don't like the look of this but presumably not related to child ages:

[paddy-r]

Also getting this, is it a leading zeroes issue?

[paddy-r]

Child ages in handovers.

1. 470_integer_child_ages, looks reasonable.

2. development, looks very similar.

[paddy-r]

@RobertClay all looks good, happy to merge if you are, although if you want a spot check of some child ages, I'd need to delve into the code a bit more.

[paddy-r]

@RobertClay don't merge this yet, got a few changes to make...

[paddy-r]

All looking good following some harmonisation with other improvements on 369-refactor.

[paddy-r]

@RobertClay Potential issue: can you check Ageing.update_binary_child_ages. I think it needs a groupby on hidp to correctly update nkids and child_ages. Currently updating at individual level, not hh level.

[RobertClay]

Hi,

Should be as simple as

• use groupby to take the first household under each hidp.
• apply child age update.
• transform back onto main vivarium population frame. use the new age chain and hidps as a dictionary.

maybe need to do same with fertility here as well. doing only once per house but make sure to account for multiple births twins two mothers etc.

not sure if you're asking me to do it or how to do it. happy to discuss more.

[RobertClay]

now I think about it theres a few things this could be done for. housing neighbourhood etc. reducing number of computations. it gets tricky though as multiple family members may be predicted different states.