problem.setcbitervar
problem.setcbitervar |
Purpose
Set a user callback to be called after each column has been tested for convergence
Synopsis
problem.setcbitervar(callback, object)
retval = callback(my_prob, my_object, colindex)
Arguments
|
callback
|
The function to be called after each column has been tested for convergence.
callback returns an integer value. The return value is interpreted as a convergence status. The possible values are:
|
||||||||
|
my_prob
|
The problem passed to the callback function.
|
||||||||
|
my_object
|
The user-defined object passed as
object to
setcbitervar.
|
||||||||
|
ColIndex
|
The number of the column which has been tested for convergence.
|
||||||||
|
object
|
A user-defined object, which can be used for any purpose by the function.
object is passed to
callback as
my_object.
|
Example
The following example sets up a callback to be executed after each variable has been tested for convergence. The user object
Important is an integer array which has already been set up and holds a flag for each variable indicating whether it is important that it converges.
Obj = None p.setcbitervar(CBIterVar, Obj)
The following sample callback function tests if the variable is already converged. If not, then it checks if the variable is important. If it is not important, the function returns a convergence status of 99.
def CBIterVar(MyProb, Obj, iCol):
(a,b,c,d,e,f,g,h,i,converged,j,k,l,m,n) = MyProb.getvar(iCol)
if converged:
return 0
if Obj[iCol]:
return 99
return -1
The
object argument is used here to hold the address of the array
Important.
Further information
This callback can be used after each variable has been checked for convergence, and allows the convergence status to be reset if required.
Related topics