1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
|
# Copyright (C) 2019 EDF
# All Rights Reserved
# This code is published under the GNU Lesser General Public License (GNU LGPL)
import numpy as np
import StOptGrids
import StOptTree
import StOptGlobal
import StOptGeners
def DynamicProgrammingByTreeHighLevel(p_grid, p_optimize, p_funcFinalValue, p_pointStock, p_initialRegime, p_fileToDump) :
# from the optimizer get back the simulation
simulator = p_optimize.getSimulator()
# final values
fin = StOptGlobal.FinalStepDP(p_grid, p_optimize.getNbRegime())
valuesNext = fin.set(p_funcFinalValue, simulator.getNodes())
ar = StOptGeners.BinaryFileArchive(p_fileToDump, "w")
nameAr = "Continuation"
nsteps =simulator.getNbStep()
# iterate on time steps
for iStep in range(nsteps) :
simulator.stepBackward()
# get back probability
proba = simulator.getProba()
# and connection matrix
connected = simulator.getConnected()
# creta tree for conditional expectation
tree=StOptTree.Tree(proba,connected)
# transition object
transStep = StOptGlobal.TransitionStepTreeDP(p_grid, p_grid, p_optimize)
valuesAndControl = transStep.oneStep(valuesNext, tree)
transStep.dumpContinuationValues(ar, nameAr, iStep, valuesNext, valuesAndControl[1], tree)
valuesNext = valuesAndControl[0]
# interpolate at the initial stock point and initial regime
return (p_grid.createInterpolator(p_pointStock).applyVec(valuesNext[p_initialRegime])).mean()
|
