search

samuel-book / stroke_outcome

Outcome modelling
MIT License
4 stars 0 forks source link

Figure formatting for mRS distribution derivations #14

Open aselaws opened 2 years ago

aselaws commented 2 years ago

I've made some plots to show how we derive the mRS distributions (yay!). Unfortunately, they're minging (boo!).

mRS_datasets_full_derivation_eurgh

The formatting is a bit nuts at the moment because I've been messing around, but you can see the sort of things that I can change - the bar colours and hatching, where the legends are and what's in them, which bars we bother to show...

Which bits of formatting do you find useful and which bits are confusing?

aselaws commented 2 years ago

Some thoughts:

  1. The effect of excess death is always small, so it's difficult to see what's going on in those bars.
  2. Maybe the horizontal stripes look like the bars are broken up into vertical sections. (Do we even need the two dists to have different hatching?)
  3. All legends apply to all axes so they need to go somewhere obvious. This might require making the figure enormous.
  4. The x-axes have an extra tick for the target value (sum of the weighted dist, or target mRS<=1 value - for the lower axis it's 0.20 and so doesn't stand out). The extra tick is liable to clash with the existing ones.
  5. The bottom axis connecting lines should go.
aselaws commented 2 years ago
  1. Should change the curved arrows to }--{ arrows that more accurately show the parts of the bars that move between axes
aselaws commented 2 years ago
  1. Can the right-hand axis y-labels move to the right-hand-side?