| 1 |
# -*- coding: utf-8 -*- |
| 2 |
|
| 3 |
######################################################## |
| 4 |
# |
| 5 |
# Copyright (c) 2003-2010 by University of Queensland |
| 6 |
# Earth Systems Science Computational Center (ESSCC) |
| 7 |
# http://www.uq.edu.au/esscc |
| 8 |
# |
| 9 |
# Primary Business: Queensland, Australia |
| 10 |
# Licensed under the Open Software License version 3.0 |
| 11 |
# http://www.opensource.org/licenses/osl-3.0.php |
| 12 |
# |
| 13 |
######################################################## |
| 14 |
|
| 15 |
__copyright__="""Copyright (c) 2003-2010 by University of Queensland |
| 16 |
Earth Systems Science Computational Center (ESSCC) |
| 17 |
http://www.uq.edu.au/esscc |
| 18 |
Primary Business: Queensland, Australia""" |
| 19 |
__license__="""Licensed under the Open Software License version 3.0 |
| 20 |
http://www.opensource.org/licenses/osl-3.0.php""" |
| 21 |
__url__="https://launchpad.net/escript-finley" |
| 22 |
|
| 23 |
""" |
| 24 |
Test suite for AMG |
| 25 |
|
| 26 |
:remark: |
| 27 |
|
| 28 |
:var __author__: name of author |
| 29 |
:var __licence__: licence agreement |
| 30 |
:var __url__: url entry point on documentation |
| 31 |
:var __version__: version |
| 32 |
:var __date__: date of the version |
| 33 |
""" |
| 34 |
|
| 35 |
__author__="Lutz Gross, l.gross@uq.edu.au" |
| 36 |
|
| 37 |
import unittest, sys |
| 38 |
|
| 39 |
from esys.escript import * |
| 40 |
from esys.finley import Rectangle,Brick |
| 41 |
from esys.escript.linearPDEs import LinearPDE, SolverOptions |
| 42 |
import numpy |
| 43 |
OPTIMIZE=True # and False |
| 44 |
SOLVER_VERBOSE=True and False |
| 45 |
|
| 46 |
MIN_MATRIX_SIZE=1 |
| 47 |
MIN_SPARSITY=1. |
| 48 |
MIN_MATRIX_SIZE=None |
| 49 |
MIN_SPARSITY=None |
| 50 |
MAX_LEVEL=None |
| 51 |
USE_AMG=True or False |
| 52 |
|
| 53 |
try: |
| 54 |
FINLEY_TEST_DATA=os.environ['FINLEY_TEST_DATA'] |
| 55 |
except KeyError: |
| 56 |
FINLEY_TEST_DATA='.' |
| 57 |
FINLEY_TEST_MESH_PATH=os.path.join(FINLEY_TEST_DATA,"data_meshes") |
| 58 |
|
| 59 |
# number of elements in the spatial directions |
| 60 |
NE_TOTAL=4096 |
| 61 |
#NE_TOTAL=4 |
| 62 |
|
| 63 |
class Test_AMG(unittest.TestCase): |
| 64 |
|
| 65 |
def test_Poisson(self): |
| 66 |
x=Solution(self.domain).getX() |
| 67 |
# --- set exact solution ---- |
| 68 |
u_ex=Scalar(1,Solution(self.domain)) |
| 69 |
g_ex=Vector(0.,Solution(self.domain)) |
| 70 |
for i in xrange(self.domain.getDim()): |
| 71 |
u_ex+=(i+1)*x[i] |
| 72 |
g_ex[i]=(i+1) |
| 73 |
|
| 74 |
# create PDE: |
| 75 |
pde=LinearPDE(self.domain,numEquations=1) |
| 76 |
pde.setSymmetryOn() |
| 77 |
mask=whereZero(x[0]) |
| 78 |
pde.setValue(r=u_ex,q=mask) |
| 79 |
pde.setValue(A=kronecker(self.domain),y=inner(g_ex,self.domain.getNormal())) |
| 80 |
# -------- get the solution --------------------------- |
| 81 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 82 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 83 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 84 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 85 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 86 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 87 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 88 |
|
| 89 |
u=pde.getSolution() |
| 90 |
# -------- test the solution --------------------------- |
| 91 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 92 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 93 |
|
| 94 |
def test_PoissonWithDirectInterpolation(self): |
| 95 |
x=Solution(self.domain).getX() |
| 96 |
# --- set exact solution ---- |
| 97 |
u_ex=Scalar(1,Solution(self.domain)) |
| 98 |
g_ex=Vector(0.,Solution(self.domain)) |
| 99 |
for i in xrange(self.domain.getDim()): |
| 100 |
u_ex+=(i+1)*x[i] |
| 101 |
g_ex[i]=(i+1) |
| 102 |
|
| 103 |
# create PDE: |
| 104 |
pde=LinearPDE(self.domain,numEquations=1) |
| 105 |
pde.setSymmetryOn() |
| 106 |
mask=whereZero(x[0]) |
| 107 |
pde.setValue(r=u_ex,q=mask) |
| 108 |
pde.setValue(A=kronecker(self.domain),y=inner(g_ex,self.domain.getNormal())) |
| 109 |
# -------- get the solution --------------------------- |
| 110 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 111 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 112 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 113 |
pde.getSolverOptions().setNumPreSweeps(3) |
| 114 |
pde.getSolverOptions().setNumPostSweeps(3) |
| 115 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 116 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 117 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 118 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 119 |
pde.getSolverOptions().setAMGInterpolation(pde.getSolverOptions().DIRECT_INTERPOLATION) |
| 120 |
|
| 121 |
u=pde.getSolution() |
| 122 |
# -------- test the solution --------------------------- |
| 123 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 124 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 125 |
|
| 126 |
def test_PoissonClassic(self): |
| 127 |
x=Solution(self.domain).getX() |
| 128 |
# --- set exact solution ---- |
| 129 |
u_ex=Scalar(1,Solution(self.domain)) |
| 130 |
g_ex=Vector(0.,Solution(self.domain)) |
| 131 |
for i in xrange(self.domain.getDim()): |
| 132 |
u_ex+=(i+1)*x[i] |
| 133 |
g_ex[i]=(i+1) |
| 134 |
|
| 135 |
# create PDE: |
| 136 |
pde=LinearPDE(self.domain,numEquations=1) |
| 137 |
pde.setSymmetryOn() |
| 138 |
mask=whereZero(x[0]) |
| 139 |
pde.setValue(r=u_ex,q=mask) |
| 140 |
pde.setValue(A=kronecker(self.domain),y=inner(g_ex,self.domain.getNormal())) |
| 141 |
# -------- get the solution --------------------------- |
| 142 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 143 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 144 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 145 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 146 |
pde.getSolverOptions().setNumPreSweeps(3) |
| 147 |
pde.getSolverOptions().setNumPostSweeps(3) |
| 148 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 149 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 150 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 151 |
pde.getSolverOptions().setAMGInterpolation(pde.getSolverOptions().CLASSIC_INTERPOLATION) |
| 152 |
|
| 153 |
u=pde.getSolution() |
| 154 |
# -------- test the solution --------------------------- |
| 155 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 156 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 157 |
|
| 158 |
def test_PoissonClassicWithFFCoupling(self): |
| 159 |
x=Solution(self.domain).getX() |
| 160 |
# --- set exact solution ---- |
| 161 |
u_ex=Scalar(1,Solution(self.domain)) |
| 162 |
g_ex=Vector(0.,Solution(self.domain)) |
| 163 |
for i in xrange(self.domain.getDim()): |
| 164 |
u_ex+=(i+1)*x[i] |
| 165 |
g_ex[i]=(i+1) |
| 166 |
|
| 167 |
# create PDE: |
| 168 |
pde=LinearPDE(self.domain,numEquations=1) |
| 169 |
pde.setSymmetryOn() |
| 170 |
mask=whereZero(x[0]) |
| 171 |
pde.setValue(r=u_ex,q=mask) |
| 172 |
pde.setValue(A=kronecker(self.domain),y=inner(g_ex,self.domain.getNormal())) |
| 173 |
# -------- get the solution --------------------------- |
| 174 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 175 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 176 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 177 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 178 |
pde.getSolverOptions().setNumPreSweeps(3) |
| 179 |
pde.getSolverOptions().setNumPostSweeps(3) |
| 180 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 181 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 182 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 183 |
pde.getSolverOptions().setAMGInterpolation(pde.getSolverOptions().CLASSIC_INTERPOLATION_WITH_FF_COUPLING) |
| 184 |
|
| 185 |
u=pde.getSolution() |
| 186 |
# -------- test the solution --------------------------- |
| 187 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 188 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 189 |
|
| 190 |
def test_PoissonSqueezedX(self): |
| 191 |
x=self.domain.getX().copy() |
| 192 |
x[0]*=0.5 |
| 193 |
self.domain.setX(x) |
| 194 |
x=Solution(self.domain).getX() |
| 195 |
# --- set exact solution ---- |
| 196 |
u_ex=Scalar(1,Solution(self.domain)) |
| 197 |
g_ex=Vector(0.,Solution(self.domain)) |
| 198 |
for i in xrange(self.domain.getDim()): |
| 199 |
u_ex+=(i+1)*x[i] |
| 200 |
g_ex[i]=(i+1) |
| 201 |
|
| 202 |
# create PDE: |
| 203 |
pde=LinearPDE(self.domain,numEquations=1) |
| 204 |
pde.setSymmetryOn() |
| 205 |
mask=whereZero(x[0]) |
| 206 |
pde.setValue(r=u_ex,q=mask) |
| 207 |
pde.setValue(A=kronecker(self.domain),y=inner(g_ex,self.domain.getNormal())) |
| 208 |
# -------- get the solution --------------------------- |
| 209 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 210 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 211 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 212 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 213 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 214 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 215 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 216 |
|
| 217 |
u=pde.getSolution() |
| 218 |
# -------- test the solution --------------------------- |
| 219 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 220 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 221 |
|
| 222 |
|
| 223 |
def test_Poisson2(self): |
| 224 |
x=Solution(self.domain).getX() |
| 225 |
# --- set exact solution ---- |
| 226 |
u_ex=Data(1.,(2,),Solution(self.domain)) |
| 227 |
g_ex=Data(0.,(2,self.domain.getDim()), Solution(self.domain)) |
| 228 |
A=Data(0.,(2,self.domain.getDim(),2,self.domain.getDim()), Function(self.domain)) |
| 229 |
for i in xrange(self.domain.getDim()): |
| 230 |
u_ex[0]+= 1*(i+1) *x[i] |
| 231 |
g_ex[0,i]=1*(i+1) |
| 232 |
u_ex[1]+= 2*(i+1)*x[i] |
| 233 |
g_ex[1,i]=2*(i+1) |
| 234 |
A[0,i,0,i]=1 |
| 235 |
A[1,i,1,i]=1 |
| 236 |
|
| 237 |
# create PDE: |
| 238 |
pde=LinearPDE(self.domain,numEquations=2) |
| 239 |
pde.setSymmetryOn() |
| 240 |
mask=whereZero(x[0])*[1,1] |
| 241 |
pde.setValue(r=u_ex,q=mask) |
| 242 |
pde.setValue(A=A,y=matrixmult(g_ex,self.domain.getNormal())) |
| 243 |
# -------- get the solution --------------------------- |
| 244 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 245 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 246 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 247 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 248 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 249 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 250 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 251 |
|
| 252 |
u=pde.getSolution() |
| 253 |
# -------- test the solution --------------------------- |
| 254 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 255 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 256 |
|
| 257 |
def test_Poisson3(self): |
| 258 |
x=Solution(self.domain).getX() |
| 259 |
# --- set exact solution ---- |
| 260 |
u_ex=Data(1.,(3,),Solution(self.domain)) |
| 261 |
g_ex=Data(0.,(3,self.domain.getDim()), Solution(self.domain)) |
| 262 |
A=Data(0.,(3,self.domain.getDim(),3,self.domain.getDim()), Function(self.domain)) |
| 263 |
for i in xrange(self.domain.getDim()): |
| 264 |
u_ex[0]+= 1*(i+1) *x[i] |
| 265 |
g_ex[0,i]=1*(i+1) |
| 266 |
u_ex[1]+= 2*(i+1)*x[i] |
| 267 |
g_ex[1,i]=2*(i+1) |
| 268 |
u_ex[2]+= 3*(i+1)*x[i] |
| 269 |
g_ex[2,i]=3*(i+1) |
| 270 |
A[0,i,0,i]=1 |
| 271 |
A[1,i,1,i]=1 |
| 272 |
A[2,i,2,i]=1 |
| 273 |
|
| 274 |
# create PDE: |
| 275 |
pde=LinearPDE(self.domain,numEquations=3) |
| 276 |
pde.setSymmetryOn() |
| 277 |
mask=whereZero(x[0])*[1,1,1] |
| 278 |
pde.setValue(r=u_ex,q=mask) |
| 279 |
pde.setValue(A=A,y=matrixmult(g_ex,self.domain.getNormal())) |
| 280 |
# -------- get the solution --------------------------- |
| 281 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 282 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 283 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 284 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 285 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 286 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 287 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 288 |
|
| 289 |
u=pde.getSolution() |
| 290 |
# -------- test the solution --------------------------- |
| 291 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 292 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 293 |
|
| 294 |
def test_Poisson4(self): |
| 295 |
x=Solution(self.domain).getX() |
| 296 |
# --- set exact solution ---- |
| 297 |
u_ex=Data(1.,(4,),Solution(self.domain)) |
| 298 |
g_ex=Data(0.,(4,self.domain.getDim()), Solution(self.domain)) |
| 299 |
A=Data(0.,(4,self.domain.getDim(),4,self.domain.getDim()), Function(self.domain)) |
| 300 |
for i in xrange(self.domain.getDim()): |
| 301 |
u_ex[0]+= 1*(i+1) *x[i] |
| 302 |
g_ex[0,i]=1*(i+1) |
| 303 |
u_ex[1]+= 2*(i+1)*x[i] |
| 304 |
g_ex[1,i]=2*(i+1) |
| 305 |
u_ex[2]+= 3*(i+1)*x[i] |
| 306 |
g_ex[2,i]=3*(i+1) |
| 307 |
u_ex[3]+= 4*(i+1)*x[i] |
| 308 |
g_ex[3,i]=4*(i+1) |
| 309 |
A[0,i,0,i]=1 |
| 310 |
A[1,i,1,i]=1 |
| 311 |
A[2,i,2,i]=1 |
| 312 |
A[3,i,3,i]=1 |
| 313 |
|
| 314 |
# create PDE: |
| 315 |
pde=LinearPDE(self.domain,numEquations=4) |
| 316 |
pde.setSymmetryOn() |
| 317 |
mask=whereZero(x[0])*[1,1,1,1] |
| 318 |
pde.setValue(r=u_ex,q=mask) |
| 319 |
pde.setValue(A=A,y=matrixmult(g_ex,self.domain.getNormal())) |
| 320 |
# -------- get the solution --------------------------- |
| 321 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 322 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 323 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 324 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 325 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 326 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 327 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 328 |
|
| 329 |
u=pde.getSolution() |
| 330 |
# -------- test the solution --------------------------- |
| 331 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 332 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 333 |
def test_Poisson2Classic(self): |
| 334 |
x=Solution(self.domain).getX() |
| 335 |
# --- set exact solution ---- |
| 336 |
u_ex=Data(1.,(2,),Solution(self.domain)) |
| 337 |
g_ex=Data(0.,(2,self.domain.getDim()), Solution(self.domain)) |
| 338 |
A=Data(0.,(2,self.domain.getDim(),2,self.domain.getDim()), Function(self.domain)) |
| 339 |
for i in xrange(self.domain.getDim()): |
| 340 |
u_ex[0]+= 1*(i+1) *x[i] |
| 341 |
g_ex[0,i]=1*(i+1) |
| 342 |
u_ex[1]+= 2*(i+1)*x[i] |
| 343 |
g_ex[1,i]=2*(i+1) |
| 344 |
A[0,i,0,i]=1 |
| 345 |
A[1,i,1,i]=1 |
| 346 |
|
| 347 |
# create PDE: |
| 348 |
pde=LinearPDE(self.domain,numEquations=2) |
| 349 |
pde.setSymmetryOn() |
| 350 |
mask=whereZero(x[0])*[1,1] |
| 351 |
pde.setValue(r=u_ex,q=mask) |
| 352 |
pde.setValue(A=A,y=matrixmult(g_ex,self.domain.getNormal())) |
| 353 |
# -------- get the solution --------------------------- |
| 354 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 355 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 356 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 357 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 358 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 359 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 360 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 361 |
pde.getSolverOptions().setAMGInterpolation(pde.getSolverOptions().CLASSIC_INTERPOLATION) |
| 362 |
|
| 363 |
u=pde.getSolution() |
| 364 |
# -------- test the solution --------------------------- |
| 365 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 366 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 367 |
|
| 368 |
def test_Poisson3Classic(self): |
| 369 |
x=Solution(self.domain).getX() |
| 370 |
# --- set exact solution ---- |
| 371 |
u_ex=Data(1.,(3,),Solution(self.domain)) |
| 372 |
g_ex=Data(0.,(3,self.domain.getDim()), Solution(self.domain)) |
| 373 |
A=Data(0.,(3,self.domain.getDim(),3,self.domain.getDim()), Function(self.domain)) |
| 374 |
for i in xrange(self.domain.getDim()): |
| 375 |
u_ex[0]+= 1*(i+1) *x[i] |
| 376 |
g_ex[0,i]=1*(i+1) |
| 377 |
u_ex[1]+= 2*(i+1)*x[i] |
| 378 |
g_ex[1,i]=2*(i+1) |
| 379 |
u_ex[2]+= 3*(i+1)*x[i] |
| 380 |
g_ex[2,i]=3*(i+1) |
| 381 |
A[0,i,0,i]=1 |
| 382 |
A[1,i,1,i]=1 |
| 383 |
A[2,i,2,i]=1 |
| 384 |
|
| 385 |
# create PDE: |
| 386 |
pde=LinearPDE(self.domain,numEquations=3) |
| 387 |
pde.setSymmetryOn() |
| 388 |
mask=whereZero(x[0])*[1,1,1] |
| 389 |
pde.setValue(r=u_ex,q=mask) |
| 390 |
pde.setValue(A=A,y=matrixmult(g_ex,self.domain.getNormal())) |
| 391 |
# -------- get the solution --------------------------- |
| 392 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 393 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 394 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 395 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 396 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 397 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 398 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 399 |
pde.getSolverOptions().setAMGInterpolation(pde.getSolverOptions().CLASSIC_INTERPOLATION) |
| 400 |
|
| 401 |
u=pde.getSolution() |
| 402 |
# -------- test the solution --------------------------- |
| 403 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 404 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 405 |
|
| 406 |
def test_Poisson4Classic(self): |
| 407 |
x=Solution(self.domain).getX() |
| 408 |
# --- set exact solution ---- |
| 409 |
u_ex=Data(1.,(4,),Solution(self.domain)) |
| 410 |
g_ex=Data(0.,(4,self.domain.getDim()), Solution(self.domain)) |
| 411 |
A=Data(0.,(4,self.domain.getDim(),4,self.domain.getDim()), Function(self.domain)) |
| 412 |
for i in xrange(self.domain.getDim()): |
| 413 |
u_ex[0]+= 1*(i+1) *x[i] |
| 414 |
g_ex[0,i]=1*(i+1) |
| 415 |
u_ex[1]+= 2*(i+1)*x[i] |
| 416 |
g_ex[1,i]=2*(i+1) |
| 417 |
u_ex[2]+= 3*(i+1)*x[i] |
| 418 |
g_ex[2,i]=3*(i+1) |
| 419 |
u_ex[3]+= 4*(i+1)*x[i] |
| 420 |
g_ex[3,i]=4*(i+1) |
| 421 |
A[0,i,0,i]=1 |
| 422 |
A[1,i,1,i]=1 |
| 423 |
A[2,i,2,i]=1 |
| 424 |
A[3,i,3,i]=1 |
| 425 |
|
| 426 |
# create PDE: |
| 427 |
pde=LinearPDE(self.domain,numEquations=4) |
| 428 |
pde.setSymmetryOn() |
| 429 |
mask=whereZero(x[0])*[1,1,1,1] |
| 430 |
pde.setValue(r=u_ex,q=mask) |
| 431 |
pde.setValue(A=A,y=matrixmult(g_ex,self.domain.getNormal())) |
| 432 |
# -------- get the solution --------------------------- |
| 433 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 434 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 435 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 436 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 437 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 438 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 439 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 440 |
pde.getSolverOptions().setAMGInterpolation(pde.getSolverOptions().CLASSIC_INTERPOLATION) |
| 441 |
|
| 442 |
u=pde.getSolution() |
| 443 |
# -------- test the solution --------------------------- |
| 444 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 445 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 446 |
|
| 447 |
def test_WeakCoupled2(self): |
| 448 |
x=Solution(self.domain).getX() |
| 449 |
# --- set exact solution ---- |
| 450 |
u_ex=Data(1.,(2,),Solution(self.domain)) |
| 451 |
g_ex=Data(0.,(2,self.domain.getDim()), Solution(self.domain)) |
| 452 |
A=Data(0.,(2,self.domain.getDim(),2,self.domain.getDim()), Function(self.domain)) |
| 453 |
for i in xrange(self.domain.getDim()): |
| 454 |
u_ex[0]+= 1*(i+1)*x[i] |
| 455 |
g_ex[0,i]=1*(i+1) |
| 456 |
u_ex[1]+= 2*(i+1)*x[i] |
| 457 |
g_ex[1,i]=2*(i+1) |
| 458 |
A[0,i,0,i]=1 |
| 459 |
A[1,i,1,i]=1 |
| 460 |
|
| 461 |
Y=Data(0.,(2,),Function(self.domain)) |
| 462 |
a=-1./2.*0.01 |
| 463 |
Y[0]= u_ex[0] + a * u_ex[1] |
| 464 |
Y[1]=a * u_ex[0] + u_ex[1] |
| 465 |
# create PDE: |
| 466 |
pde=LinearPDE(self.domain,numEquations=2) |
| 467 |
pde.setSymmetryOn() |
| 468 |
mask=whereZero(x[0])*[1,1] |
| 469 |
pde.setValue(r=u_ex,q=mask) |
| 470 |
pde.setValue(A=A, |
| 471 |
D=kronecker(2)+a*(numpy.ones((2,2))-kronecker(2)), |
| 472 |
Y=Y, |
| 473 |
y=matrixmult(g_ex,self.domain.getNormal())) |
| 474 |
# -------- get the solution --------------------------- |
| 475 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 476 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 477 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 478 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 479 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 480 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 481 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 482 |
|
| 483 |
u=pde.getSolution() |
| 484 |
# -------- test the solution --------------------------- |
| 485 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 486 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 487 |
|
| 488 |
def test_WeakCoupled3(self): |
| 489 |
x=Solution(self.domain).getX() |
| 490 |
# --- set exact solution ---- |
| 491 |
u_ex=Data(1.,(3,),Solution(self.domain)) |
| 492 |
g_ex=Data(0.,(3,self.domain.getDim()), Solution(self.domain)) |
| 493 |
A=Data(0.,(3,self.domain.getDim(),3,self.domain.getDim()), Function(self.domain)) |
| 494 |
for i in xrange(self.domain.getDim()): |
| 495 |
u_ex[0]+= 1*(i+1)*x[i] |
| 496 |
g_ex[0,i]=1*(i+1) |
| 497 |
u_ex[1]+= 2*(i+1)*x[i] |
| 498 |
g_ex[1,i]=2*(i+1) |
| 499 |
u_ex[2]+= 3*(i+1)*x[i] |
| 500 |
g_ex[2,i]=3*(i+1) |
| 501 |
A[0,i,0,i]=1 |
| 502 |
A[1,i,1,i]=1 |
| 503 |
A[2,i,2,i]=1 |
| 504 |
|
| 505 |
a=-1./3.*0.01 |
| 506 |
Y=Data(0.,(3,),Function(self.domain)) |
| 507 |
Y[0]= u_ex[0]+ a * u_ex[1]+ a * u_ex[2] |
| 508 |
Y[1]= a * u_ex[0]+ u_ex[1]+ a * u_ex[2] |
| 509 |
Y[2]= a * u_ex[0]+ a * u_ex[1]+ u_ex[2] |
| 510 |
# create PDE: |
| 511 |
pde=LinearPDE(self.domain,numEquations=3) |
| 512 |
pde.setSymmetryOn() |
| 513 |
mask=whereZero(x[0])*[1,1,1] |
| 514 |
pde.setValue(r=u_ex,q=mask) |
| 515 |
pde.setValue(A=A, |
| 516 |
D=kronecker(3)+a*(numpy.ones((3,3))-kronecker(3)), |
| 517 |
Y=Y, |
| 518 |
y=matrixmult(g_ex,self.domain.getNormal())) |
| 519 |
# -------- get the solution --------------------------- |
| 520 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 521 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 522 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 523 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 524 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 525 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 526 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 527 |
u=pde.getSolution() |
| 528 |
# -------- test the solution --------------------------- |
| 529 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 530 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 531 |
|
| 532 |
def test_WeakCoupled4(self): |
| 533 |
x=Solution(self.domain).getX() |
| 534 |
# --- set exact solution ---- |
| 535 |
u_ex=Data(1.,(4,),Solution(self.domain)) |
| 536 |
g_ex=Data(0.,(4,self.domain.getDim()), Solution(self.domain)) |
| 537 |
A=Data(0.,(4,self.domain.getDim(),4,self.domain.getDim()), Function(self.domain)) |
| 538 |
for i in xrange(self.domain.getDim()): |
| 539 |
u_ex[0]+= 1*(i+1)*x[i] |
| 540 |
g_ex[0,i]=1*(i+1) |
| 541 |
u_ex[1]+= 2*(i+1)*x[i] |
| 542 |
g_ex[1,i]=2*(i+1) |
| 543 |
u_ex[2]+= 3*(i+1)*x[i] |
| 544 |
g_ex[2,i]=3*(i+1) |
| 545 |
u_ex[3]+= 4*(i+1)*x[i] |
| 546 |
g_ex[3,i]=4*(i+1) |
| 547 |
A[0,i,0,i]=1 |
| 548 |
A[1,i,1,i]=1 |
| 549 |
A[2,i,2,i]=1 |
| 550 |
A[3,i,3,i]=1 |
| 551 |
|
| 552 |
a=-1./4.*0.01 |
| 553 |
|
| 554 |
Y=Data(0.,(4,),Function(self.domain)) |
| 555 |
Y[0]= u_ex[0]+ a * u_ex[1]+ a * u_ex[2]+ a * u_ex[3] |
| 556 |
Y[1]= a * u_ex[0]+ u_ex[1]+ a * u_ex[2]+ a * u_ex[3] |
| 557 |
Y[2]= a * u_ex[0]+ a * u_ex[1]+ u_ex[2]+ a * u_ex[3] |
| 558 |
Y[3]= a * u_ex[0]+ a * u_ex[1]+ a * u_ex[2]+ u_ex[3] |
| 559 |
# create PDE: |
| 560 |
pde=LinearPDE(self.domain,numEquations=4) |
| 561 |
pde.setSymmetryOn() |
| 562 |
mask=whereZero(x[0])*[1,1,1,1] |
| 563 |
pde.setValue(r=u_ex,q=mask) |
| 564 |
pde.setValue(A=A, |
| 565 |
D=kronecker(4)+a*(numpy.ones((4,4))-kronecker(4)), |
| 566 |
Y=Y, |
| 567 |
y=matrixmult(g_ex,self.domain.getNormal())) |
| 568 |
# -------- get the solution --------------------------- |
| 569 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 570 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 571 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 572 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 573 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 574 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 575 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 576 |
u=pde.getSolution() |
| 577 |
# -------- test the solution --------------------------- |
| 578 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 579 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 580 |
|
| 581 |
def test_StrongCoupled2(self): |
| 582 |
x=Solution(self.domain).getX() |
| 583 |
# --- set exact solution ---- |
| 584 |
u_ex=Data(1.,(2,),Solution(self.domain)) |
| 585 |
g_ex=Data(0.,(2,self.domain.getDim()), Solution(self.domain)) |
| 586 |
A=Data(0.,(2,self.domain.getDim(),2,self.domain.getDim()), Function(self.domain)) |
| 587 |
for i in xrange(self.domain.getDim()): |
| 588 |
u_ex[0]+= 1*(i+1)*x[i] |
| 589 |
g_ex[0,i]=1*(i+1) |
| 590 |
u_ex[1]+= 2*(i+1)*x[i] |
| 591 |
g_ex[1,i]=2*(i+1) |
| 592 |
A[0,i,0,i]=1 |
| 593 |
A[1,i,1,i]=1 |
| 594 |
|
| 595 |
Y=Data(0.,(2,),Function(self.domain)) |
| 596 |
a=-1./2.*0.99 |
| 597 |
Y[0]= u_ex[0] + a * u_ex[1] |
| 598 |
Y[1]=a * u_ex[0] + u_ex[1] |
| 599 |
# create PDE: |
| 600 |
pde=LinearPDE(self.domain,numEquations=2) |
| 601 |
pde.setSymmetryOn() |
| 602 |
mask=whereZero(x[0])*[1,1] |
| 603 |
pde.setValue(r=u_ex,q=mask) |
| 604 |
pde.setValue(A=A, |
| 605 |
D=kronecker(2)+a*(numpy.ones((2,2))-kronecker(2)), |
| 606 |
Y=Y, |
| 607 |
y=matrixmult(g_ex,self.domain.getNormal())) |
| 608 |
# -------- get the solution --------------------------- |
| 609 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 610 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 611 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 612 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 613 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 614 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 615 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 616 |
|
| 617 |
u=pde.getSolution() |
| 618 |
# -------- test the solution --------------------------- |
| 619 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 620 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 621 |
|
| 622 |
def test_StrongCoupled3(self): |
| 623 |
x=Solution(self.domain).getX() |
| 624 |
# --- set exact solution ---- |
| 625 |
u_ex=Data(1.,(3,),Solution(self.domain)) |
| 626 |
g_ex=Data(0.,(3,self.domain.getDim()), Solution(self.domain)) |
| 627 |
A=Data(0.,(3,self.domain.getDim(),3,self.domain.getDim()), Function(self.domain)) |
| 628 |
for i in xrange(self.domain.getDim()): |
| 629 |
u_ex[0]+= 1*(i+1)*x[i] |
| 630 |
g_ex[0,i]=1*(i+1) |
| 631 |
u_ex[1]+= 2*(i+1)*x[i] |
| 632 |
g_ex[1,i]=2*(i+1) |
| 633 |
u_ex[2]+= 3*(i+1)*x[i] |
| 634 |
g_ex[2,i]=3*(i+1) |
| 635 |
A[0,i,0,i]=1 |
| 636 |
A[1,i,1,i]=1 |
| 637 |
A[2,i,2,i]=1 |
| 638 |
|
| 639 |
a=-1./3.*0.99 |
| 640 |
Y=Data(0.,(3,),Function(self.domain)) |
| 641 |
Y[0]= u_ex[0]+ a * u_ex[1]+ a * u_ex[2] |
| 642 |
Y[1]= a * u_ex[0]+ u_ex[1]+ a * u_ex[2] |
| 643 |
Y[2]= a * u_ex[0]+ a * u_ex[1]+ u_ex[2] |
| 644 |
# create PDE: |
| 645 |
pde=LinearPDE(self.domain,numEquations=3) |
| 646 |
pde.setSymmetryOn() |
| 647 |
mask=whereZero(x[0])*[1,1,1] |
| 648 |
pde.setValue(r=u_ex,q=mask) |
| 649 |
pde.setValue(A=A, |
| 650 |
D=kronecker(3)+a*(numpy.ones((3,3))-kronecker(3)), |
| 651 |
Y=Y, |
| 652 |
y=matrixmult(g_ex,self.domain.getNormal())) |
| 653 |
# -------- get the solution --------------------------- |
| 654 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 655 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 656 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 657 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 658 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 659 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 660 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 661 |
u=pde.getSolution() |
| 662 |
# -------- test the solution --------------------------- |
| 663 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 664 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 665 |
|
| 666 |
def test_StrongCoupled4(self): |
| 667 |
x=Solution(self.domain).getX() |
| 668 |
# --- set exact solution ---- |
| 669 |
u_ex=Data(1.,(4,),Solution(self.domain)) |
| 670 |
g_ex=Data(0.,(4,self.domain.getDim()), Solution(self.domain)) |
| 671 |
A=Data(0.,(4,self.domain.getDim(),4,self.domain.getDim()), Function(self.domain)) |
| 672 |
for i in xrange(self.domain.getDim()): |
| 673 |
u_ex[0]+= 1*(i+1)*x[i] |
| 674 |
g_ex[0,i]=1*(i+1) |
| 675 |
u_ex[1]+= 2*(i+1)*x[i] |
| 676 |
g_ex[1,i]=2*(i+1) |
| 677 |
u_ex[2]+= 3*(i+1)*x[i] |
| 678 |
g_ex[2,i]=3*(i+1) |
| 679 |
u_ex[3]+= 4*(i+1)*x[i] |
| 680 |
g_ex[3,i]=4*(i+1) |
| 681 |
A[0,i,0,i]=1 |
| 682 |
A[1,i,1,i]=1 |
| 683 |
A[2,i,2,i]=1 |
| 684 |
A[3,i,3,i]=1 |
| 685 |
|
| 686 |
a=-1./4.*0.99 |
| 687 |
|
| 688 |
Y=Data(0.,(4,),Function(self.domain)) |
| 689 |
Y[0]= u_ex[0]+ a * u_ex[1]+ a * u_ex[2]+ a * u_ex[3] |
| 690 |
Y[1]= a * u_ex[0]+ u_ex[1]+ a * u_ex[2]+ a * u_ex[3] |
| 691 |
Y[2]= a * u_ex[0]+ a * u_ex[1]+ u_ex[2]+ a * u_ex[3] |
| 692 |
Y[3]= a * u_ex[0]+ a * u_ex[1]+ a * u_ex[2]+ u_ex[3] |
| 693 |
# create PDE: |
| 694 |
pde=LinearPDE(self.domain,numEquations=4) |
| 695 |
pde.setSymmetryOn() |
| 696 |
mask=whereZero(x[0])*[1,1,1,1] |
| 697 |
pde.setValue(r=u_ex,q=mask) |
| 698 |
pde.setValue(A=A, |
| 699 |
D=kronecker(4)+a*(numpy.ones((4,4))-kronecker(4)), |
| 700 |
Y=Y, |
| 701 |
y=matrixmult(g_ex,self.domain.getNormal())) |
| 702 |
# -------- get the solution --------------------------- |
| 703 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 704 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 705 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 706 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 707 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 708 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 709 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 710 |
u=pde.getSolution() |
| 711 |
# -------- test the solution --------------------------- |
| 712 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 713 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 714 |
|
| 715 |
|
| 716 |
def test_Square(self): |
| 717 |
# PDE constants |
| 718 |
h1 = 0.5 |
| 719 |
h2 = 0.5 |
| 720 |
beta = 1. |
| 721 |
alpha1 = 2. |
| 722 |
alpha2 = 1. |
| 723 |
|
| 724 |
# domain masks and domain-specific constants |
| 725 |
x = Solution(self.domain).getX(); x0 = x[0]; x1 = x[1] |
| 726 |
omega2 = wherePositive(x0-h1)*wherePositive(x1-h2) |
| 727 |
omega1 = 1-omega2 |
| 728 |
ratio = alpha1/alpha2 |
| 729 |
alpha = alpha1*omega1 + alpha2*omega2 |
| 730 |
|
| 731 |
# --- set exact solution ---- |
| 732 |
a1 = 1. |
| 733 |
d1 = 1. |
| 734 |
b1 = -d1*h2 |
| 735 |
c1 = -d1*h1 |
| 736 |
|
| 737 |
a2 = a1 - (1.-ratio)*d1*h1*h2 |
| 738 |
b2 = ratio*b1 |
| 739 |
c2 = ratio*c1 |
| 740 |
d2 = ratio*d1 |
| 741 |
|
| 742 |
u_ex = omega1*(a1 + b1*x0 + c1*x1 + d1*x0*x1) + \ |
| 743 |
omega2*(a2 + b2*x0 + c2*x1 + d2*x0*x1) |
| 744 |
|
| 745 |
# create PDE: |
| 746 |
pde = LinearPDE(self.domain,numEquations=1) |
| 747 |
|
| 748 |
# set the value to that of the solution on the boundary |
| 749 |
q = whereZero(x0) + whereZero(x1) + \ |
| 750 |
whereZero(sup(x0)-x0) + whereZero(sup(x1)-x1) |
| 751 |
pde.setValue(q=q,r=u_ex) |
| 752 |
|
| 753 |
# create X points in the centre of the grid elements |
| 754 |
xe = Function(self.domain).getX() |
| 755 |
x0 = xe[0] |
| 756 |
x1 = xe[1] |
| 757 |
|
| 758 |
# redefine omega so that apha is more precise on the diagonal (?) |
| 759 |
omega2 = wherePositive(x0-h1)*wherePositive(x1-h2) |
| 760 |
omega1 = 1-omega2 |
| 761 |
ratio = alpha1/alpha2 |
| 762 |
alpha = alpha1*omega1 + alpha2*omega2 |
| 763 |
|
| 764 |
# set up PDE coefficients |
| 765 |
pde.setValue(A=alpha*kronecker(self.domain), D=beta, Y=beta*u_ex) |
| 766 |
pde.setSymmetryOn() |
| 767 |
|
| 768 |
# -------- get the solution --------------------------- |
| 769 |
pde.getSolverOptions().setTolerance(self.SOLVER_TOL) |
| 770 |
pde.getSolverOptions().setSolverMethod(SolverOptions.PCG) |
| 771 |
if (USE_AMG): pde.getSolverOptions().setPreconditioner(SolverOptions.AMG) |
| 772 |
pde.getSolverOptions().setVerbosity(SOLVER_VERBOSE) |
| 773 |
if MIN_MATRIX_SIZE!= None: pde.getSolverOptions().setMinCoarseMatrixSize(MIN_MATRIX_SIZE) |
| 774 |
if MIN_SPARSITY!=None: pde.getSolverOptions().setMinCoarseMatrixSparsity(MIN_SPARSITY) |
| 775 |
if MAX_LEVEL!=None: pde.getSolverOptions().setLevelMax(MAX_LEVEL) |
| 776 |
u = pde.getSolution() |
| 777 |
|
| 778 |
# -------- test the solution --------------------------- |
| 779 |
error=Lsup(u-u_ex)/Lsup(u_ex) |
| 780 |
self.assertTrue(error<self.RES_TOL, "solution error %s is too big."%error) |
| 781 |
|
| 782 |
|
| 783 |
class Test_AMGOnFinleyHex2DOrder1(Test_AMG): |
| 784 |
RES_TOL=5.e-7 |
| 785 |
SOLVER_TOL=1.e-8 |
| 786 |
def setUp(self): |
| 787 |
NE=int(float(NE_TOTAL)**(1./2.)+0.5) |
| 788 |
self.domain = Rectangle(NE,NE,1, optimize=OPTIMIZE) |
| 789 |
def tearDown(self): |
| 790 |
del self.domain |
| 791 |
|
| 792 |
class Test_AMGOnFinleyHex3DOrder1(Test_AMG): |
| 793 |
RES_TOL=5.e-7 |
| 794 |
SOLVER_TOL=1.e-8 |
| 795 |
def setUp(self): |
| 796 |
NE=int(float(NE_TOTAL)**(1./3.)+0.5) |
| 797 |
self.domain = Brick(NE,NE,NE,1, optimize=OPTIMIZE) |
| 798 |
def tearDown(self): |
| 799 |
del self.domain |
| 800 |
if __name__ == '__main__': |
| 801 |
suite = unittest.TestSuite() |
| 802 |
suite.addTest(unittest.makeSuite(Test_AMGOnFinleyHex2DOrder1)) |
| 803 |
suite.addTest(unittest.makeSuite(Test_AMGOnFinleyHex3DOrder1)) |
| 804 |
# suite.addTest(Test_AMGOnFinleyHex3DOrder1("test_Poisson4")) |
| 805 |
# suite.addTest(Test_AMGOnFinleyHex2DOrder1("test_WeakCoupled4")) |
| 806 |
|
| 807 |
s=unittest.TextTestRunner(verbosity=2).run(suite) |
| 808 |
if not s.wasSuccessful(): sys.exit(12) |
Parent Directory
| 
Revision Log