seabbs
commented
2 years ago Nice.
- I agree on using bars vs arrows (I just had no idea how to do this).
- I like the dashed/non-dashed lines. I changed to truncated delay as that is obviously exactly what it is but perhaps it is a bit on the nose.
- For the other subjects I have tried to include a sample of the other kinds of censoring we might see and so probably want to discuss.. So:
- Wider than a day censoring
- No censoring for primary event
- No censoring for secondary event
- To make this more schematic like I've dropped the y axis line but see what you think.
We might also want to think about including the interaction between censoring and truncation on this schematic (i.e having a observation time on B and having the censoring window overlap with it).
I've merged for now but not implying this is final.
Not sure this usefully improves things and I ran out of arrow making energy before wrapping up.