diff --git a/docs/About/Acknowledgements.html b/docs/About/Acknowledgements.html index b0e5239d..85d17ed9 100644 --- a/docs/About/Acknowledgements.html +++ b/docs/About/Acknowledgements.html @@ -19,13 +19,13 @@
MyModel
and MyCode
MyModel
and MyCode
MyModel
must supply a replacement for Utility(). Since utility depends on the current state, the method must be automatic (not static). Here is an example with one state variable and one action and how they might determine utility.
#import "DDP" @@ -78,7 +78,7 @@
As explained elsewhere, if
-s
contains a state variable its "value" is not simply themselves. Likewisea
. Their current values are retrieved by sending them to CV(). Also, note that \(U()\) at a state is always treated as a vector-valued function in DDP . SoCV(d)
is a column vector. As a state variables
is a scalar at \(\theta\).
z*
Indicator vectors make it possible to write any utility as a sum: \(U = \sum_k d_k u_k\). But as with the single action approach the interpretation is not obvious and \(U()\) will in no way resemble usual mathematical notation for an objective.
-
+
- Example:
decl exper = new ActionCounter("Yrs Experience",10,work); //track up to 10 years working
EndogenousStates(exper);
-
Order + | Order | +number of lags [default = 1] |
wrked = new LaggedAction("Worked Last Year",work);+
wrked = new LaggedAction("Worked Last Year",work); +
prevoccup = new LaggedState("Prev",occup);
+prevoccup = new LaggedState("Prev",occup); +
retired = new PermanentChoice("Ret",retire);
+retired = new PermanentChoice("Ret",retire); +@@ -10478,10 +10486,12 @@
+ s = new StateVariable("s",5); s->MakeTerminal(<3;4>); v = new StateVariable("v",1); v->MakeTerminal(1); +
CV(s)=3
or CV(s)=4
or CV(v)=1
diff --git a/docs/FiveO/Algorithms.ox.html b/docs/FiveO/Algorithms.ox.html
index 705e0a63..a45f709a 100644
--- a/docs/FiveO/Algorithms.ox.html
+++ b/docs/FiveO/Algorithms.ox.html
@@ -22,7 +22,7 @@ Modelbase
aspects of parameters are stored separately, at least from the user's perspective. Parameters can be fixed or free, but other kinds of constraints and relationships between parameters are left to the user. A key difference with FiveO is that parameters are represented as objects derived from built-in parameter types. Parameters are created using a class that captures the constraints on the parameter and its relation to other parameters. For example, with one line of code a user can add a vector of parameters that are guaranteed to be Increasing during optimization: x1 >x2 > ยทยทยท > xM.
-