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
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
|
#!/usr/bin/python3
# Copyright (C) 2016 EDF
# All Rights Reserved
# This code is published under the GNU Lesser General Public License (GNU LGPL)
import math
import numpy as np
import unittest
import StOptGrids
import StOptReg
import StOptGlobal
import Utils
import Simulators as sim
import Optimizers as opt
import dp.DynamicProgrammingByRegressionDist as dynmpi
import dp.SimulateRegressionDist as srtmpi
import importlib
accuracyClose = 1.5
# p_ndim dimension of the swing
# p_bOneFile Do we use one unique file for continuation values
def swingND(p_ndim, p_bOneFile):
from mpi4py import MPI
world = MPI.COMM_WORLD
initialValues = np.zeros(1) + 1.
sigma = np.zeros(1) + 0.2
mu = np.zeros(1) + 0.05
corr = np.ones((1,1))
# number of step
nStep = 20
# exercise date
dates = np.linspace(0., 1., nStep + 1)
N = 3 # number of exercises
T = 1.
strike = 1.
nbSimul = 80000
nMesh = 8
# payoff for swing
payOffBasket = Utils.BasketCall(strike)
payoff = Utils.PayOffFictitiousSwing(payOffBasket,N)
nbMesh = np.zeros(1, dtype = np.int32) + nMesh
# simulator
simulator = sim.BlackScholesSimulator(initialValues, sigma, mu, corr, dates[len(dates) - 1], len(dates) - 1, nbSimul, False)
# Grid
lowValues = np.zeros(p_ndim)
step = np.zeros(p_ndim) + 1.
# the stock is discretized with values from 0 to N included
nbStep = np.zeros(p_ndim, dtype = np.int32) + N
grid = StOptGrids.RegularSpaceGrid(lowValues, step, nbStep)
# optimizer
optimizer = opt.OptimizerFictitiousSwingBlackScholes(payOffBasket, N, p_ndim)
# initial values
initialStock = np.zeros(p_ndim)
initialRegime = 0
fileToDump = "CondExpSwingOptimNDHLMpi"
# regressor
regressor = StOptReg.LocalLinearRegression(nbMesh)
# link the simulations to the optimizer
optimizer.setSimulator(simulator)
# bermudean value
valueOptimMpi = dynmpi.DynamicProgrammingByRegressionDist(grid, optimizer, regressor, payoff, initialStock, initialRegime, fileToDump, p_bOneFile)
# simulation value
simulatorForward = sim.BlackScholesSimulator(initialValues, sigma, mu, corr, dates[len(dates) - 1], len(dates) - 1, nbSimul, True)
optimizer.setSimulator(simulatorForward)
valSimuMpi = srtmpi.SimulateRegressionDist(grid, optimizer, payoff, initialStock, initialRegime, fileToDump, p_bOneFile)
print("Optim", valueOptimMpi, "SIMU", valSimuMpi)
return valueOptimMpi, valSimuMpi
class testSwingOptimSimuNDMpiTest(unittest.TestCase):
def test_swingOptionOptim2DSimuMpiOneFile(self):
moduleMpi4Py=importlib.util.find_spec('mpi4py')
if (moduleMpi4Py is not None):
from mpi4py import MPI
world = MPI.COMM_WORLD
ndim = 2
bOneFile = True
val = swingND(ndim, bOneFile)
if world.rank == 0:
self.assertAlmostEqual(val[0], val[1], None, None, accuracyClose)
def test_swingOptionOptim2DSimuMpiMultipleFile(self):
moduleMpi4Py=importlib.util.find_spec('mpi4py')
if (moduleMpi4Py is not None):
from mpi4py import MPI
world = MPI.COMM_WORLD
ndim = 2
bOneFile = False
val = swingND(ndim, bOneFile)
if world.rank == 0:
self.assertAlmostEqual(val[0], val[1], None, None, accuracyClose)
if __name__ == '__main__':
unittest.main()
|
