See #1169 for details. This PR changes regimen::ii() to just just update the dose interval, with no other fancy tricks. Look at the working example in #1169; I think that's what we want.
I put a note in the tests that I took the cutoffs for those tests from looking at a plot of the profile. The plot looks right to me. The tests check that the pattern we get is what we'd expect from the model code. But the exact cutoffs aren't obvious.
@kyleam I'd like to send this through as a hotfix if possible. As for general reliability, I've been kicking the tires for several hours on this working up a blog post and no other surprises found.
See #1169 for details. This PR changes
regimen::ii()to just just update the dose interval, with no other fancy tricks. Look at the working example in #1169; I think that's what we want.I put a note in the tests that I took the cutoffs for those tests from looking at a plot of the profile. The plot looks right to me. The tests check that the pattern we get is what we'd expect from the model code. But the exact cutoffs aren't obvious.
@kyleam I'd like to send this through as a hotfix if possible. As for general reliability, I've been kicking the tires for several hours on this working up a blog post and no other surprises found.