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

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

481

482

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

548

549

550

551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

576

577

578

579

580

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

625

626

627

628

629

630

631

632

633

634

635

636

637

638

639

640

641

642

643

644

645

646

647

648

649

650

651

652

653

654

655

656

657

658

659

660

661

662

663

664

665

666

667

668

669

670

671

672

673

674

675

676

677

678

679

680

681

682

683

684

685

686

687

688

689

690

691

692

693

694

695

696

# https://pyrocko.org - GPLv3 

# 

# The Pyrocko Developers, 21st Century 

# ---|P------/S----------~Lg---------- 

import numpy as num 

import logging 

 

from pyrocko import moment_tensor as mt 

import pyrocko.guts as guts 

from pyrocko.guts import Float, String, Timestamp, Int 

from pyrocko.model import Location 

from pyrocko.modelling import okada_ext 

from pyrocko.util import get_threadpool_limits 

 

guts_prefix = 'modelling' 

 

logger = logging.getLogger('pyrocko.modelling.okada') 

 

d2r = num.pi/180. 

r2d = 180./num.pi 

km = 1e3 

 

 

class AnalyticalSource(Location): 

name = String.T( 

optional=True, 

default='') 

 

time = Timestamp.T( 

default=0., 

help='source origin time', 

optional=True) 

 

vr = Float.T( 

default=0., 

help='Rupture velocity', 

optional=True) 

 

@property 

def northing(self): 

return self.north_shift 

 

@property 

def easting(self): 

return self.east_shift 

 

clone = guts.clone 

 

 

class AnalyticalRectangularSource(AnalyticalSource): 

''' 

Rectangular analytical source model 

''' 

 

strike = Float.T( 

default=0.0, 

help='strike direction in [deg], measured clockwise from north') 

 

dip = Float.T( 

default=90.0, 

help='dip angle in [deg], measured downward from horizontal') 

 

rake = Float.T( 

default=0.0, 

help='rake angle in [deg], ' 

'measured counter-clockwise from right-horizontal ' 

'in on-plane view') 

 

al1 = Float.T( 

default=0., 

help='Distance "left" side to source point [m]') 

 

al2 = Float.T( 

default=0., 

help='Distance "right" side to source point [m]') 

 

aw1 = Float.T( 

default=0., 

help='Distance "lower" side to source point [m]') 

 

aw2 = Float.T( 

default=0., 

help='Distance "upper" side to source point [m]') 

 

slip = Float.T( 

default=0., 

help='Slip on the rectangular source area [m]', 

optional=True) 

 

@property 

def length(self): 

return abs(-self.al1 + self.al2) 

 

@property 

def width(self): 

return abs(-self.aw1 + self.aw2) 

 

@property 

def area(self): 

return self.width * self.length 

 

 

class OkadaSource(AnalyticalRectangularSource): 

''' 

Rectangular Okada source model 

''' 

 

opening = Float.T( 

default=0., 

help='Opening of the plane in [m]', 

optional=True) 

 

poisson = Float.T( 

default=0.25, 

help='Poisson\'s ratio, typically 0.25', 

optional=True) 

 

shearmod = Float.T( 

default=32e9, 

help='Shear modulus along the plane [Pa]', 

optional=True) 

 

@property 

def lamb(self): 

''' 

Calculation of first Lame's parameter 

 

According to Mueller (2007), the first Lame parameter lambda can be 

determined from the formulation for the poisson ration :math:`\\nu`: 

:math:`\\nu = \\frac{\\lambda}{2(\\lambda + \\mu)}` 

with the shear modulus :math:`\\mu` 

''' 

 

return (2. * self.poisson * self.shearmod) / (1. - 2*self.poisson) 

 

@property 

def seismic_moment(self): 

''' 

Scalar Seismic moment 

 

Code copied from Kite 

Disregarding the opening (as for now) 

We assume a shear modulus of :math:`\\mu = 36 \\mathrm{GPa}` 

and :math:`M_0 = mu A D` 

 

.. important :: 

 

We assume a perfect elastic solid with :math:`K=\\frac{5}{3}\\mu` 

 

Through :math:`\\mu = \\frac{3K(1-2\\nu)}{2(1+\\nu)}` this leads to 

:math:`\\mu = \\frac{8(1+\\nu)}{1-2\\nu}` 

 

:return: Seismic moment release 

:rtype: float 

''' 

 

if self.shearmod: 

mu = self.shearmod 

elif self.poisson: 

self.shearmod = (8. * (1. + self.poisson)) / (1. - 2*self.poisson) 

mu = self.shearmod 

else: 

raise ValueError( 

'Shear modulus or poisson ratio needed for moment calculation') 

 

disl = 0. 

if self.slip: 

disl = (disl**2 + self.slip**2)**.5 

if self.opening: 

disl = (disl**2 + self.opening**2)**.5 

 

return mu * self.area * disl 

 

@property 

def moment_magnitude(self): 

''' Moment magnitude from Seismic moment 

 

We assume :math:`M_\\mathrm{w} = {\\frac{2}{3}}\\log_{10}(M_0) - 10.7` 

 

:returns: Moment magnitude 

:rtype: float 

''' 

return mt.moment_to_magnitude(self.seismic_moment) 

 

def source_patch(self): 

''' Build source information array for okada_ext.okada input 

 

:return: array of the source data as input for okada_ext.okada 

:rtype: :py:class:`numpy.ndarray`, ``(1, 9)`` 

''' 

return num.array([ 

self.northing, 

self.easting, 

self.depth, 

self.strike, 

self.dip, 

self.al1, 

self.al2, 

self.aw1, 

self.aw2]) 

 

def source_disloc(self): 

''' Build source dislocation for okada_ext.okada input 

 

:return: array of the source dislocation data as input for 

okada_ext.okada 

:rtype: :py:class:`numpy.ndarray`, ``(1, 3)`` 

''' 

return num.array([ 

num.cos(self.rake * d2r) * self.slip, 

num.sin(self.rake * d2r) * self.slip, 

self.opening]) 

 

def discretize(self, nlength, nwidth, *args, **kwargs): 

''' Discretize the given fault by ``nlength * nwidth`` fault patches 

 

Discretizing the fault into several sub faults. ``nlength`` is 

number of points in strike direction, ``nwidth`` in down dip direction 

along the fault. Fault orientation, slip and elastic parameters are 

kept. 

 

:param nlength: Number of discrete points in faults strike direction 

:type nlength: int 

:param nwidth: Number of discrete points in faults down-dip direction 

:type nwidth: int 

:return: Discrete fault patches 

:rtype: list of :py:class:`pyrocko.modelling.okada.OkadaSource` objects 

''' 

assert nlength > 0 

assert nwidth > 0 

 

il = num.repeat(num.arange(nlength), nwidth) 

iw = num.tile(num.arange(nwidth), nlength) 

 

patch_length = self.length / nlength 

patch_width = self.width / nwidth 

 

al1 = -patch_length / 2. 

al2 = patch_length / 2. 

aw1 = -patch_width / 2. 

aw2 = patch_width / 2. 

 

source_points = num.zeros((nlength * nwidth, 3)) 

source_points[:, 0] = il * patch_length + patch_length / 2. 

source_points[:, 1] = iw * patch_width + patch_width / 2. 

 

source_points[:, 0] += self.al1 

source_points[:, 1] -= self.aw2 

 

rotmat = num.asarray( 

mt.euler_to_matrix(self.dip*d2r, self.strike*d2r, 0.)) 

 

source_points_rot = num.dot(rotmat.T, source_points.T).T 

source_points_rot[:, 0] += self.northing 

source_points_rot[:, 1] += self.easting 

source_points_rot[:, 2] += self.depth 

 

kwargs = { 

prop: getattr(self, prop) for prop in self.T.propnames 

if prop not in [ 

'north_shift', 'east_shift', 'depth', 

'al1', 'al2', 'aw1', 'aw2']} 

 

return ( 

[OkadaPatch( 

parent=self, 

ix=src_point[0], 

iy=src_point[1], 

north_shift=coord[0], 

east_shift=coord[1], 

depth=coord[2], 

al1=al1, al2=al2, aw1=aw1, aw2=aw2, **kwargs) 

for src_point, coord in zip(source_points, source_points_rot)], 

source_points) 

 

 

class OkadaPatch(OkadaSource): 

 

ix = Int.T(help='Relative index of the patch in x') 

iy = Int.T(help='Relative index of the patch in y') 

 

def __init__(self, *args, parent=None, **kwargs): 

super().__init__(*args, **kwargs) 

self.parent = parent 

 

 

class DislocationInverter(object): 

''' 

Toolbox for Boundary Element Method (BEM) and dislocation inversion based 

on okada_ext.okada 

''' 

 

@staticmethod 

def get_coef_mat(source_patches_list, pure_shear=False, 

rotate_sdn=True, nthreads=1): 

''' 

Build coefficient matrix for given source_patches 

 

The BEM for a fault and the determination of the slip distribution from 

the stress drop is based on the relation 

:math:`stress = coefmat \\cdot displ`. 

Here the coefficient matrix is build and filled based on the 

okada_ext.okada displacements and partial displacement 

differentiations. 

 

:param source_patches_list: list of all OkadaSources, which shall be 

used for BEM 

:type source_patches_list: list of 

:py:class:`pyrocko.modelling.okada.OkadaSource` 

:param pure_shear: Shall only shear forces be taken into account, 

or additionally include opening, default False. 

:type pure_shear: optional, bool 

:param rotate_sdn: Rotation towards strike, dip, normal, default True. 

:type rotate_sdn: optional, bool 

:param nthreads: Number of threads, default 1 

:type nthreads: optional, int 

 

:return: coefficient matrix for all sources 

:rtype: :py:class:`numpy.ndarray`, 

``(len(source_patches_list) * 3, len(source_patches_list) * 3)`` 

''' 

 

source_patches = num.array([ 

src.source_patch() for src in source_patches_list]) 

receiver_coords = source_patches[:, :3].copy() 

 

npoints = len(source_patches_list) 

 

if pure_shear: 

n_eq = 2 

else: 

n_eq = 3 

 

coefmat = num.zeros((npoints * 3, npoints * 3)) 

 

lambda_mean = num.mean([src.lamb for src in source_patches_list]) 

mu_mean = num.mean([src.shearmod for src in source_patches_list]) 

 

unit_disl = 1. 

disl_cases = { 

'strikeslip': { 

'slip': unit_disl, 

'opening': 0., 

'rake': 0.}, 

'dipslip': { 

'slip': unit_disl, 

'opening': 0., 

'rake': 90.}, 

'tensileslip': { 

'slip': 0., 

'opening': unit_disl, 

'rake': 0.} 

} 

 

diag_ind = [0, 4, 8] 

kron = num.zeros(9) 

kron[diag_ind] = 1. 

kron = kron[num.newaxis, num.newaxis, :] 

 

for idisl, case_type in enumerate([ 

'strikeslip', 'dipslip', 'tensileslip'][:n_eq]): 

case = disl_cases[case_type] 

source_disl = num.array([ 

case['slip'] * num.cos(case['rake'] * d2r), 

case['slip'] * num.sin(case['rake'] * d2r), 

case['opening']]) 

 

results = okada_ext.okada( 

source_patches, 

num.tile(source_disl, npoints).reshape(-1, 3), 

receiver_coords, 

lambda_mean, 

mu_mean, 

nthreads=nthreads, 

rotate_sdn=rotate_sdn, 

stack_sources=False) 

 

eps = 0.5 * (results[:, :, 3:] + 

results[:, :, (3, 6, 9, 4, 7, 10, 5, 8, 11)]) 

 

dilatation = eps[:, :, diag_ind].sum(axis=-1)[:, :, num.newaxis] 

 

stress_sdn = kron*lambda_mean*dilatation + 2.*mu_mean*eps 

coefmat[:, idisl::3] = stress_sdn[:, :, (2, 5, 8)]\ 

.reshape(-1, npoints*3).T 

 

if pure_shear: 

coefmat[2::3, :] = 0. 

 

return -coefmat / unit_disl 

 

@staticmethod 

def get_coef_mat_single(source_patches_list, pure_shear=False, 

rotate_sdn=True, nthreads=1): 

''' 

Build coefficient matrix for given source_patches 

 

The BEM for a fault and the determination of the slip distribution from 

the stress drop is based on the relation 

:math:`stress = coefmat \\cdot displ`. 

Here the coefficient matrix is build and filled based on the 

okada_ext.okada displacements and partial displacement 

differentiations. 

 

:param source_patches_list: list of all OkadaSources, which shall be 

used for BEM 

:type source_patches_list: list of 

:py:class:`pyrocko.modelling.okada.OkadaSource` 

:param pure_shear: Flag, if also opening mode shall be taken into 

account (False) or the fault is described as pure shear (True). 

:type pure_shear: optional, bool 

:param rotate_sdn: Rotation towards strike, dip, normal, default True. 

:type rotate_sdn: optional, bool 

:param nthreads: Number of threads, default 1 

:type nthreads: optional, int 

 

:return: coefficient matrix for all sources 

:rtype: :py:class:`numpy.ndarray`, 

``(len(source_patches_list) * 3, len(source_patches_list) * 3)`` 

''' 

 

source_patches = num.array([ 

src.source_patch() for src in source_patches_list]) 

receiver_coords = source_patches[:, :3].copy() 

 

npoints = len(source_patches_list) 

 

if pure_shear: 

n_eq = 2 

else: 

n_eq = 3 

 

coefmat = num.zeros((npoints * 3, npoints * 3)) 

 

lambda_mean = num.mean([src.lamb for src in source_patches_list]) 

mu_mean = num.mean([src.shearmod for src in source_patches_list]) 

 

unit_disl = 1. 

disl_cases = { 

'strikeslip': { 

'slip': unit_disl, 

'opening': 0., 

'rake': 0.}, 

'dipslip': { 

'slip': unit_disl, 

'opening': 0., 

'rake': 90.}, 

'tensileslip': { 

'slip': 0., 

'opening': unit_disl, 

'rake': 0.} 

} 

 

diag_ind = [0, 4, 8] 

kron = num.zeros(9) 

kron[diag_ind] = 1. 

kron = kron[num.newaxis, :] 

 

for idisl, case_type in enumerate([ 

'strikeslip', 'dipslip', 'tensileslip'][:n_eq]): 

case = disl_cases[case_type] 

source_disl = num.array([ 

case['slip'] * num.cos(case['rake'] * d2r), 

case['slip'] * num.sin(case['rake'] * d2r), 

case['opening']]) 

 

for isrc, source in enumerate(source_patches): 

results = okada_ext.okada( 

source[num.newaxis, :], 

source_disl[num.newaxis, :], 

receiver_coords, 

lambda_mean, 

mu_mean, 

nthreads=nthreads, 

rotate_sdn=rotate_sdn) 

 

eps = \ 

0.5 * ( 

results[:, 3:] + 

results[:, (3, 6, 9, 4, 7, 10, 5, 8, 11)]) 

 

dilatation = num.sum(eps[:, diag_ind], axis=1)[:, num.newaxis] 

stress_sdn = kron * lambda_mean * dilatation+2. * mu_mean * eps 

 

coefmat[:, isrc*3 + idisl] = stress_sdn[:, (2, 5, 8)].ravel() 

 

if pure_shear: 

coefmat[2::3, isrc * 3 + idisl] = 0. 

 

return -coefmat / unit_disl 

 

@staticmethod 

def get_coef_mat_slow(source_patches_list, pure_shear=False, 

rotate_sdn=True, nthreads=1): 

''' 

Build coefficient matrix for given source_patches (Slow version) 

 

The BEM for a fault and the determination of the slip distribution from 

the stress drop is based on the relation 

:math:`stress = coefmat \\cdot displ`. 

Here the coefficient matrix is build and filled based on the 

okada_ext.okada displacements and partial displacement 

differentiations. 

 

:param source_patches_list: list of all OkadaSources, which shall be 

used for BEM 

:type source_patches_list: list of 

:py:class:`pyrocko.modelling.okada.OkadaSource` 

:param pure_shear: Flag, if also opening mode shall be taken into 

account (False) or the fault is described as pure shear (True). 

:type pure_shear: optional, bool 

:param rotate_sdn: Rotation towards strike, dip, normal, default True. 

:type rotate_sdn: optional, bool 

:param nthreads: Number of threads, default 1 

:type nthreads: optional, int 

 

:return: coefficient matrix for all sources 

:rtype: :py:class:`numpy.ndarray`, 

``(source_patches_list.shape[0] * 3, 

source_patches_list.shape[0] * 3(2))`` 

''' 

 

source_patches = num.array([ 

src.source_patch() for src in source_patches_list]) 

receiver_coords = source_patches[:, :3].copy() 

 

npoints = len(source_patches_list) 

 

if pure_shear: 

n_eq = 2 

else: 

n_eq = 3 

 

coefmat = num.zeros((npoints * 3, npoints * 3)) 

 

def ned2sdn_rotmat(strike, dip): 

rotmat = mt.euler_to_matrix( 

(dip + 180.) * d2r, strike * d2r, 0.).A 

return rotmat 

 

lambda_mean = num.mean([src.lamb for src in source_patches_list]) 

shearmod_mean = num.mean([src.shearmod for src in source_patches_list]) 

 

unit_disl = 1. 

disl_cases = { 

'strikeslip': { 

'slip': unit_disl, 

'opening': 0., 

'rake': 0.}, 

'dipslip': { 

'slip': unit_disl, 

'opening': 0., 

'rake': 90.}, 

'tensileslip': { 

'slip': 0., 

'opening': unit_disl, 

'rake': 0.} 

} 

for idisl, case_type in enumerate([ 

'strikeslip', 'dipslip', 'tensileslip'][:n_eq]): 

case = disl_cases[case_type] 

source_disl = num.array([ 

case['slip'] * num.cos(case['rake'] * d2r), 

case['slip'] * num.sin(case['rake'] * d2r), 

case['opening']]) 

 

for isource, source in enumerate(source_patches): 

results = okada_ext.okada( 

source[num.newaxis, :].copy(), 

source_disl[num.newaxis, :].copy(), 

receiver_coords, 

lambda_mean, 

shearmod_mean, 

nthreads=nthreads, 

rotate_sdn=rotate_sdn) 

 

for irec in range(receiver_coords.shape[0]): 

eps = num.zeros((3, 3)) 

for m in range(3): 

for n in range(3): 

eps[m, n] = 0.5 * ( 

results[irec][m * 3 + n + 3] + 

results[irec][n * 3 + m + 3]) 

 

stress = num.zeros((3, 3)) 

dilatation = num.sum([eps[i, i] for i in range(3)]) 

 

for m, n in zip([0, 0, 0, 1, 1, 2], [0, 1, 2, 1, 2, 2]): 

if m == n: 

stress[m, n] = \ 

lambda_mean * \ 

dilatation + \ 

2. * shearmod_mean * \ 

eps[m, n] 

 

else: 

stress[m, n] = \ 

2. * shearmod_mean * \ 

eps[m, n] 

stress[n, m] = stress[m, n] 

 

normal = num.array([0., 0., -1.]) 

for isig in range(3): 

tension = num.sum(stress[isig, :] * normal) 

coefmat[irec * n_eq + isig, isource * n_eq + idisl] = \ 

tension / unit_disl 

 

return coefmat 

 

@staticmethod 

def get_disloc_lsq( 

stress_field, 

coef_mat=None, 

source_list=None, 

pure_shear=False, 

epsilon=None, 

nthreads=1, 

**kwargs): 

''' 

Least square inversion to get displacement from stress 

 

Follows approach for Least-Square Inversion published in Menke (1989) 

to calculate displacements on a fault with several segments from a 

given stress field. If not done, the coefficient matrix is determined 

within the code. 

 

:param stress_field: Array containing the stress change [Pa] for each 

source patch (order: [ 

src1 dstress_Strike, src1 dstress_Dip, src1 dstress_Tensile, 

src2 dstress_Strike, ...]) 

:type stress_field: :py:class:`numpy.ndarray`, ``(n_sources * 3, )`` 

:param coef_mat: Coefficient matrix to connect source patches 

displacement and the resulting stress field 

:type coef_mat: optional, :py:class:`numpy.ndarray`, 

``(source_patches_list.shape[0] * 3, 

source_patches.shape[] * 3(2)`` 

:param source_list: list of all OkadaSources, which shall be 

used for BEM 

:type source_list: optional, list of 

:py:class:`pyrocko.modelling.okada.OkadaSource` 

:param epsilon: regularize small values from the coefficient matrix, 

default None. 

:type epsilon: optional, float 

 

:param nthreads: number of threads for the least squares inversion, 

default 1 

:type nthreads: optional, int 

 

:return: inverted displacements (u_strike, u_dip , u_tensile) for each 

source patch. order: [ 

[patch1 u_Strike, patch1 u_Dip, patch1 u_Tensile], 

[patch2 u_Strike, patch2 u_Dip, patch2 u_Tensile], 

...] 

:rtype: :py:class:`numpy.ndarray`, ``(n_sources, 3)`` 

''' 

 

if source_list is not None and coef_mat is None: 

coef_mat = DislocationInverter.get_coef_mat( 

source_list, pure_shear=pure_shear, nthreads=nthreads, 

**kwargs) 

 

if epsilon is not None: 

coef_mat[coef_mat < epsilon] = 0. 

 

idx = num.arange(0, coef_mat.shape[0]) 

if pure_shear: 

idx = idx[idx % 3 != 2] 

 

coef_mat_in = coef_mat[idx, :][:, idx] 

disloc_est = num.zeros(coef_mat.shape[0]) 

 

if stress_field.ndim == 2: 

stress_field = stress_field.ravel() 

 

threadpool_limits = get_threadpool_limits() 

 

with threadpool_limits(limits=nthreads, user_api='blas'): 

try: 

disloc_est[idx] = num.linalg.multi_dot([ 

num.linalg.inv(num.dot(coef_mat_in.T, coef_mat_in)), 

coef_mat_in.T, 

stress_field[idx]]) 

except num.linalg.LinAlgError as e: 

logger.warning('Linear inversion failed!') 

logger.warning( 

'coef_mat: %s\nstress_field: %s', 

coef_mat_in, stress_field[idx]) 

raise e 

return disloc_est.reshape(-1, 3) 

 

 

__all__ = [ 

'AnalyticalSource', 

'AnalyticalRectangularSource', 

'OkadaSource', 

'DislocationInverter']