Your state variables are transiting to an unreachable state:
NOISY
.Transit()
method for the variable. Volume=NOISY
will slow the code down considerably. You should lower the volume once problems have been resolved.Using default Utility() equal to 0. Your derived DDP should provide a replacement for Bellman::Utility().
MyModel
list Utility()
as a non-static method (see all examples). Utility()
in the .ox file. If you don't do this a Ox runtime error will occur saying the routine does not exist.Setting iseed to 0 means ranseed is not reset
Number of replications R not set or invalid. Setting to 1.
The creator routine for OneDimensionalChoice states has not been called. Runtime errors likely.
Dynamic approximation to continuous state has not defined. Call SetKeep().
Using AgeBrackets with deterministic aging: consider using Aging().
Simulating ergodic paths without fixed Tmax? Path may be of unbounded length.
Masking data for observability.
print level invalid
Tracking of auxiliary outcomes still experimental. It may not work.
Do not add to tracking list after predictions made ... ignored
Transitions and actual values …
Transitions and actual values are fixed. They are computed in CreateSpaces() and never again.
Setting update time to OnlyOnce. Transitions and actual values do not depend on fixed or random effect values. If they do, results are not reliable.
Setting update time to AfterFixed. Transitions and actual values can depend on fixed effect values but not random effects. If they do, results are not reliable.
Setting update time to AfterRandom. Transitions and actual values can depend on fixed and random effects, which is safe but may be redundant and therefore slower than necessary.
Clock must be set before calling SubsampleStates. Setting clock type to InfiniteHorizon.
State Space Already Defined. DryRun request ignored.
Only a dry run of creating the state space Theta will be performed. Program ends at the end of CreateSpaces().
No actions have been added to the model. A no-choice action inserted.
Clock has not been set. Setting clock type to InfiniteHorizon.
clock is ergodic but terminal states exist? Inconsistency in the set up.
Linear systems solution for stationary distribution returns code 2. May be unreliable.
User code has called the default Solve() function for GSolve(instate)
User code has called the default Solve() function for Method. Does not do anything
User code has not set the discount factor yet. Setting it to default value of 0.90
Must call SubSampleStates() before you use KeaneWolpin::Solve().
Using KW approximazation on a model with infeasible actions at some states. All reachable states at a given time t for which the approximation is used must have the same feasible action set for results to be sensible.
Label for discrete value should be a string.
a. Trigger values must be integer or matrix b. Reset value must be integer between -1 and b.N