ImperialCollegeLondon / epidemia

epidemia package
https://imperialcollegelondon.github.io/epidemia/index.html
47 stars 15 forks source link

Include NPIs in R(t) plots as in published paper and reports #29

Closed flaxter closed 3 years ago

flaxter commented 4 years ago

@ettieunwin is going to share code with @jscott6

jscott6 commented 4 years ago

just want to make sure we actually want this feature. It would apply only to binary covariates. How would we deal with indicators that could potentially switch multiple times? Also are the provided covariate names necessarily what we want to end up with in the final plot?

An alternative would be to allow the user to manually add this as a layer to the ggplot object returned by plot_rt.

s-mishra commented 4 years ago

I think we should take labels and timings as input else this won't remain general.

jscott6 commented 4 years ago

Cool - we can do this as either a list of data frames (one for each group), or one data frame with a group column

agandy commented 4 years ago

Agreed that this should be kept as simple as possible.

jscott6 commented 3 years ago

Having thought about this - I think it's best done out of the package for now. ggplots in epidemia are easy to modify and add layers to. Will think about reopening if i find time (or if anyone else has time).