-
Notifications
You must be signed in to change notification settings - Fork 0
/
REVISION NOTES.txt
17 lines (13 loc) · 961 Bytes
/
REVISION NOTES.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
REVISION NOTES
- implemented a GoF test in CJS model
- read in AHM book that posterior predictive check can only work on aggregate measures, not on single cell values
- ran CJS model in m-array for proper GoF test -> tested with and without transients
- removed chicks from original data, but retained all other 226 individuals
- DIC is a very confusing measure and pD, the number of effective parameters is highly variable
- for prediction AIC is recommended, hence I extracted the number of parameters and calculated AIC by hand from deviance
- created model selection table and adopted most parsimonious model (phi constant, p time)
- adopted multi-event formulation for transience model
- included this model in the IPM
- reverted fecundity part of model back to a binary good/bad year formulation
- changed prior for juvenile survival to value consistent with Oro 2004 Balearic Shearwater
- added breeding propensity to be 0.85 - 0.95 annually varying