Differences

This shows you the differences between two versions of the page.

--- admb_real:all_cohorts:model_description [2010/11/16 13:55] – lennartfr
+++ admb_real:all_cohorts:model_description [2010/11/16 14:25] (current) – lennartfr
@@ Line 1: / Line 1: @@
 ===== Model following ALL cohorts =====
+=== Description ===
+Most of the models presented so far have followed one cohort in the first year under consideration, two in the second, three in the third, and so on. In addition, [[:hovedside#model_estimating_only_3_year_olds_and_older|the models following ages 3 and up]] have relied on a preprocessing of the input data in order to ignore age groups zero through two. We now present a model which follows a fixed number of cohorts every year, without the need to preprocess the data.
+The properties of the model are given in the following table:
+^ Input data (example) | start_year (1994), end_year (2001), minage (3), maxage (12). Corresponding catch (C) and survey data (I). |
+^ Population trajectory  | Matrix N, indexed as N(i,j) = cohort aged j in year i. |
+^ Unknowns to be estimated | First row and column of N, catchability q, survey standard deviation s, mortality M. |
+^ Statistical assumption | log(I) ~ N( log(qN), s). |
+^ Recurrence relation | N(i,j) = [ N(i-1,j-1) - C(i-1,j-1) ] exp(-M). |
+:!: **NOTE** that N is structured differently than for the models in the other sections. The same goes for catch and survey data in the input. Essentially, the diagonals of this N correspond to the N used in other models, and similarly for the catch matrix.
 === ADMB implementation ===