naobservatory / p2ra

4 stars 1 forks source link

Plot estimates with zero counts differently #216

Closed dp-rice closed 1 year ago

dp-rice commented 1 year ago

For estimates with no viral reads, don't plot the boxes, instead indicate that we have upper bounds on our estimates by plotting:

incidence-boxen

prevalence-boxen

Note that for the plot by location, we will still plot the boxes since these estimates use information borrowed from the other locations.

by_location_incidence-boxen

One other minor change: we now give the vertical bar for the median the same alpha value as the other elements.

dp-rice commented 1 year ago

Final version:

incidence-violin

prevalence-violin

by_location_incidence-violin