Add line plot of estimates over time for spatio-temporal model

athowes / multi-agyw

Spatio-temporal estimates of HIV risk group proportions for AGYW across 13 priority countries in sub-Saharan Africa

https://athowes.github.io/multi-agyw

MIT License

2 stars 2 forks source link

Add line plot of estimates over time for spatio-temporal model #18

Closed athowes closed 2 years ago

athowes commented 3 years ago

Coloured by age_idx
Should be an artefact of aaa_fit_all-dhs-multi-sexbehav-sae
Aim is to provide visualisation of how much temporal variation there is in the estimates
Answer relevant questions like "is behaviour getting risker" (in general, in any specific places or age groups)

athowes commented 3 years ago

This is useful for now as a exploratory plot but there isn't really a message that it's trying to convey other than "for the most part things are not changing very much"
It's OK with 5 but it would be not possible to do this with 13 countries (too many colours to tell what's going on)
I can pick out some lines which appear to be moving around a lot, so perhaps warrants investigation into what's going on there
To include this with 13 countires, would make each country one aggregated point and then plot (rather than trying to show all countries)

athowes commented 3 years ago

Showing uncertainty with alpha is nice, but this plot is obviously too busy to see anything. Not much of a story about temporal trends other than there isn't much of one.

athowes commented 2 years ago

I don't think that a plot of temporal trends need be included in the final paper. Perhaps in a Shiny dashboard