When I pass a list of random integers betweenin the interval (1,0)$(1,0)$ as function arguments, the function behaves as I expect it to.
However, when NMaximize calls the function, I get the following error:
NMaximize::nnum: The function value -1000-19 If[a[21]==1,500,1000] is not a number at {a[1],a[2],a[3],a[4],a[5],a[6],a[7],a[8],a[9],a[10],a[11],a[12],a[13],a[14],a[15],a[16],a[17],a[18],a[19],a[20]} = {1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.,1.}. >>
The interesting part of the error is the complaint about a[21]; this variable doesn't exist unless the function is called by NMaximize. The idea is that NMaximize will find integer, binary values for the parameters that will maximize the return value of the function.
NMaximize::nnum: The function value -1000 - 19 If[a[21] == 1,500, 1000] is not a number at {a[1], a[2], a[3], a[4], a[5], a[6], a[7], a[8], a[9], a[10], a[11], a[12], a[13], a[14], a[15], a[16], a[17], a[18], a[19], a[20]} = {1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.}. >>
The interesting part of the error is the complaint abouta[21]; this variable doesn't exist unless the function is called byNMaximize. The idea is thatNMaximize` will find integer, binary values for the parameters that will maximize the return value of the function.
My guess is that there is problem in how my function is being translated internally by NMaximizeNMaximize, but I am not not sure what to do at this point.
Help.
TIA,
Todd Dixon
(*Assign Initial Values*) Clear[f, i, a, vars, realconstraints, integerconstraints] PeriodCapacityLoss=10;PeriodCapacityLoss = 10; InitialCapacity = 1000; OOSCapacity=500;OOSCapacity = 500; AssymtoticCapacity=200;AssymtoticCapacity = 200; Periods = 20; CurrentCapacity[1]=InitialCapacity;CurrentCapacity[1] = InitialCapacity; (*Generate random series of cleaning flags*) For[j=1For[j = 1, j<j Periods+1< Periods + 1, j++, RecoveryFlag[j]= RecoveryFlag[j] = RandomInteger[{0, 1}]; ]; (*Function to simulate effect of capacity degredation and recovery*) f[a_] := Module[{i}, (* Set initial condition as clean *) CurrentCapacity[1] = InitialCapacity; For[i=2 For[i = 2, i<i Periods+1< Periods + 1, i++, CurrentCapacity[i] = If[a[i] == 0 && a[i - 1] == 0, CurrentCapacity[i - 1] - PeriodCapacityLoss, If[a[i] == 1, OOSCapacity, InitialCapacity] ]; ]; ]; Return[Total[Map[CurrentCapacity, Range[Periods]]]]; ]; (*Pass random cleaning flags to degredation function and plot*) Print[f[RecoveryFlag]]; ListLinePlot[Array[CurrentCapacity, Periods]] (* Define the integer constraints *) vars=Array[avars = Array[a, Periods]; realconstraints = And@@Map[Greater[2And @@ Map[Greater[2, #, 0]&, vars]; integerconstraints = Append[realconstraints, Element[vars, Integers]]; (* Find the value of the recovery flag that maximizes capacity across the time window *) NMaximize[{f[a], integerconstraints}, vars, Method->{"DifferentialEvolution"}] 
