Final changes to structure and vignettes as a result of going through everything to write the first draft of the package paper.
Structural changes:
- Removed the specific prior for
ind_y_0
as that was double-counting the first observation in the likelihood. The parameterind_y_0
is still estimated in the data, but does not get a direct prior and instead has its prior constructed through they_hat
vector as do the other sizes. - Removed
y_0_obs
as passed data because it isn't used in the model once the prior forind_y_0
was removed. - Data assignment, data generation, and testing code had to be re-worked to account for the removal of
y_0_obs
. - Added a function to plot sizes over time as well as the DE plot function.
Vignette changes:
- Added the Here Be Dragons vignette that investigates an interaction between errors in numerical methods in the longitudinal model and the MCMC sampling.
- Re-worded and re-structured other vignettes to take advantage of the new size over time function.
- Included a set of posterior estimates for the Canham demo because running the fit takes a few hours.
What's Changed
- Maths vignette by @Tess-LaCoil in #44
- Vignette fonti by @Tess-LaCoil in #46
- Here be dragons by @Tess-LaCoil in #45
- Added a function to plot sizes over time in parallel with the DE plot… by @Tess-LaCoil in #48
- Paper writing by @Tess-LaCoil in #50
Full Changelog: 1.0...1.1