pythonot.github.io/gen_modules/ot.mapping.html at master · PythonOT/pythonot.github.io

Repository navigation

History

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

697

698

699

700

701

702

703

704

705

706

707

708

709

710

711

712

713

714

715

716

717

718

719

720

721

722

723

724

725

726

727

728

729

730

731

732

733

734

735

736

737

738

739

740

741

742

743

744

745

746

747

748

749

750

751

752

753

754

755

756

757

758

759

760

761

762

763

764

765

766

767

768

769

770

771

772

773

774

775

776

777

778

779

780

781

782

783

784

785

786

787

788

789

790

791

792

793

794

795

796

797

798

799

800

801

802

803

804

805

806

807

808

809

810

811

812

813

814

815

816

817

818

819

820

821

822

823

824

825

826

827

828

829

830

831

832

833

834

835

836

837

838

839

840

841

842

843

844

845

846

847

848

849

850

851

852

853

854

<!DOCTYPE html>

<head>

<title>ot.mapping — POT Python Optimal Transport 0.9.7 documentation</title>

</head>

POT Python Optimal Transport

</a>

</form>

</div>

</div><div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="Navigation menu">

<li class="toctree-l1"><a class="reference internal" href="../index.html">POT: Python Optimal Transport</a></li>

<li class="toctree-l1"><a class="reference internal" href="../auto_examples/plot_quickstart_guide.html">Quickstart Guide</a></li>

<li class="toctree-l1"><a class="reference internal" href="../auto_examples/index.html">Examples gallery</a></li>

<li class="toctree-l1"><a class="reference internal" href="../user_guide.html">User guide</a></li>

<li class="toctree-l1 current"><a class="reference internal" href="../all.html">API and modules</a><ul class="current">

<li class="toctree-l2"><a class="reference internal" href="ot.backend.html">ot.backend</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.batch.html">ot.batch</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.bregman.html">ot.bregman</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.datasets.html">ot.datasets</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.factored.html">ot.factored</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.gaussian.html">ot.gaussian</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.gromov.html">ot.gromov</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.lowrank.html">ot.lowrank</a></li>

<li class="toctree-l2 current"><a class="current reference internal" href="#">ot.mapping</a><ul>

<li class="toctree-l3"><a class="reference internal" href="#functions">Functions</a><ul>

<li class="toctree-l4"><a class="reference internal" href="#ot.mapping.joint_OT_mapping_kernel"><code class="docutils literal notranslate">joint_OT_mapping_kernel()</code></a></li>

<li class="toctree-l4"><a class="reference internal" href="#ot.mapping.joint_OT_mapping_linear"><code class="docutils literal notranslate">joint_OT_mapping_linear()</code></a></li>

<li class="toctree-l4"><a class="reference internal" href="#ot.mapping.nearest_brenier_potential_fit"><code class="docutils literal notranslate">nearest_brenier_potential_fit()</code></a></li>

<li class="toctree-l4"><a class="reference internal" href="#examples-using-ot-mapping-nearest-brenier-potential-fit">Examples using <code class="docutils literal notranslate">ot.mapping.nearest_brenier_potential_fit</code></a></li>

<li class="toctree-l4"><a class="reference internal" href="#examples-using-ot-mapping-nearest-brenier-potential-predict-bounds">Examples using <code class="docutils literal notranslate">ot.mapping.nearest_brenier_potential_predict_bounds</code></a></li>

</ul>

</li>

<li class="toctree-l3"><a class="reference internal" href="#id0"><code class="docutils literal notranslate">joint_OT_mapping_kernel()</code></a></li>

<li class="toctree-l3"><a class="reference internal" href="#id8"><code class="docutils literal notranslate">joint_OT_mapping_linear()</code></a></li>

<li class="toctree-l3"><a class="reference internal" href="#id11"><code class="docutils literal notranslate">nearest_brenier_potential_fit()</code></a></li>

<li class="toctree-l3"><a class="reference internal" href="#id13"><code class="docutils literal notranslate">nearest_brenier_potential_predict_bounds()</code></a></li>

</ul>

</li>

<li class="toctree-l2"><a class="reference internal" href="ot.optim.html">ot.optim</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.partial.html">ot.partial</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.regpath.html">ot.regpath</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.sliced.html">ot.sliced</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.smooth.html">ot.smooth</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.stochastic.html">ot.stochastic</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.unbalanced.html">ot.unbalanced</a></li>

<li class="toctree-l2"><a class="reference internal" href="ot.utils.html">ot.utils</a></li>

<li class="toctree-l2"><a class="reference internal" href="../all.html#module-ot">Main <code class="xref py py-mod docutils literal notranslate">ot</code> functions</a></li>

</ul>

</li>

<li class="toctree-l1"><a class="reference internal" href="../releases.html">Releases</a></li>

<li class="toctree-l1"><a class="reference internal" href="../contributors.html">Contributors</a></li>

<li class="toctree-l1"><a class="reference internal" href="../contributing.html">Contributing to POT</a></li>

<li class="toctree-l1"><a class="reference internal" href="../code_of_conduct.html">Code of conduct</a></li>

</ul>

</div>

</nav>

<a href="../index.html">POT Python Optimal Transport</a>

</nav>

<li class="breadcrumb-item"><a href="../all.html">API and modules</a></li>

<li class="breadcrumb-item active">ot.mapping</li>

<a href="../_sources/gen_modules/ot.mapping.rst.txt" rel="nofollow"> View page source</a>

</li>

</ul>

<hr/>

</div>

<h1>ot.mapping<a class="headerlink" href="#module-ot.mapping" title="Link to this heading"></a></h1>

Optimal Transport maps and variants

Warning

Note that by default the module is not imported in <a class="reference internal" href="../all.html#module-ot" title="ot"><code class="xref py py-mod docutils literal notranslate">ot</code></a>. In order to

use it you need to explicitly import <a class="reference internal" href="#module-ot.mapping" title="ot.mapping"><code class="xref py py-mod docutils literal notranslate">ot.mapping</code></a>

</div>

<h2>Functions<a class="headerlink" href="#functions" title="Link to this heading"></a></h2>

ot.mapping.joint_OT_mapping_kernel(xs, xt, mu=1, eta=0.001, kerneltype='gaussian', sigma=1, bias=False, verbose=False, verbose2=False, numItermax=100, numInnerItermax=10, stopInnerThr=1e-06, stopThr=1e-05, log=False, **kwargs)<a class="reference internal" href="../_modules/ot/mapping.html#joint_OT_mapping_kernel">[source]</a><a class="headerlink" href="#ot.mapping.joint_OT_mapping_kernel" title="Link to this definition"></a></dt>

<dd>Joint OT and nonlinear mapping estimation with kernels as proposed in

[8].

The function solves the following optimization problem:

\[ \begin{align}\begin{aligned}\min_{\gamma, L\in\mathcal{H}}\quad \|L(\mathbf{X_s}) -

n_s\gamma \mathbf{X_t}\|^2_F + \mu \langle \gamma, \mathbf{M} \rangle_F +

\eta \|L\|^2_\mathcal{H}\\s.t. \ \gamma \mathbf{1} = \mathbf{a}\\ \gamma^T \mathbf{1} = \mathbf{b}\\ \gamma \geq 0\end{aligned}\end{align} \]</div>

where :

<li>\(\mathbf{M}\) is the (<cite>ns</cite>, <cite>nt</cite>) squared euclidean cost matrix between samples in

\(\mathbf{X_s}\) and \(\mathbf{X_t}\) (scaled by \(n_s\))</li>

<li>\(L\) is a \(n_s \times d\) linear operator on a kernel matrix that

approximates the barycentric mapping</li>

<li>\(\mathbf{a}\) and \(\mathbf{b}\) are uniform source and target weights</li>

</ul>

The problem consist in solving jointly an optimal transport matrix

\(\gamma\) and the nonlinear mapping that fits the barycentric mapping

\(n_s\gamma \mathbf{X_t}\).

One can also estimate a mapping with constant bias (see supplementary

material of [8]) using the bias optional argument.

The algorithm used for solving the problem is the block coordinate

descent that alternates between updates of \(\mathbf{G}\) (using conditional gradient)

and the update of \(\mathbf{L}\) using a classical kernel least square solver.

<dt class="field-odd">Parameters:</dt>

<li>xs (array-like (ns,d)) – samples in the source domain</li>

<li>xt (array-like (nt,d)) – samples in the target domain</li>

<li>mu (<a class="reference external" href="https://docs.python.org/3/library/functions.html#float" title="(in Python v3.14)">float</a>,optional) – Weight for the linear OT loss (>0)</li>

<li>eta (<a class="reference external" href="https://docs.python.org/3/library/functions.html#float" title="(in Python v3.14)">float</a>, optional) – Regularization term for the linear mapping L (>0)</li>

<li>kerneltype (<a class="reference external" href="https://docs.python.org/3/library/stdtypes.html#str" title="(in Python v3.14)">str</a>,optional) – kernel used by calling function <a class="reference internal" href="ot.utils.html#id103" title="ot.utils.kernel"><code class="xref py py-func docutils literal notranslate">ot.utils.kernel()</code></a> (gaussian by default)</li>

<li>sigma (<a class="reference external" href="https://docs.python.org/3/library/functions.html#float" title="(in Python v3.14)">float</a>, optional) – Gaussian kernel bandwidth.</li>

<li>bias (<a class="reference external" href="https://docs.python.org/3/library/functions.html#bool" title="(in Python v3.14)">bool</a>,optional) – Estimate linear mapping with constant bias</li>

<li>verbose (<a class="reference external" href="https://docs.python.org/3/library/functions.html#bool" title="(in Python v3.14)">bool</a>, optional) – Print information along iterations</li>

<li>verbose2 (<a class="reference external" href="https://docs.python.org/3/library/functions.html#bool" title="(in Python v3.14)">bool</a>, optional) – Print information along iterations</li>

<li>numItermax (<a class="reference external" href="https://docs.python.org/3/library/functions.html#int" title="(in Python v3.14)">int</a>, optional) – Max number of BCD iterations</li>

<li>numInnerItermax (<a class="reference external" href="https://docs.python.org/3/library/functions.html#int" title="(in Python v3.14)">int</a>, optional) – Max number of iterations (inner CG solver)</li>

<li>stopInnerThr (<a class="reference external" href="https://docs.python.org/3/library/functions.html#float" title="(in Python v3.14)">float</a>, optional) – Stop threshold on error (inner CG solver) (>0)</li>

<li>stopThr (<a class="reference external" href="https://docs.python.org/3/library/functions.html#float" title="(in Python v3.14)">float</a>, optional) – Stop threshold on relative loss decrease (>0)</li>

<li>log (<a class="reference external" href="https://docs.python.org/3/library/functions.html#bool" title="(in Python v3.14)">bool</a>, optional) – record log if True</li>

</ul>

</dd>

<dt class="field-even">Returns:</dt>

<li>gamma ((ns, nt) array-like) – Optimal transportation matrix for the given parameters</li>

<li>L ((ns, d) array-like) – Nonlinear mapping matrix ((\(n_s+1\), <cite>d</cite>) if bias)</li>

<li>log (dict) – log dictionary return only if log==True in parameters</li>

</ul>

</dd>

</dl>

References

[8]

M. Perrot, N. Courty, R. Flamary, A. Habrard,

“Mapping estimation for discrete optimal transport”,

Neural Information Processing Systems (NIPS), 2016.

</aside>

See also

<dt><a class="reference internal" href="ot.lp.html#ot.lp.emd" title="ot.lp.emd"><code class="xref py py-obj docutils literal notranslate">ot.lp.emd</code></a></dt><dd>Unregularized OT

</dd>

<dt><a class="reference internal" href="ot.optim.html#id0" title="ot.optim.cg"><code class="xref py py-obj docutils literal notranslate">ot.optim.cg</code></a></dt><dd>General regularized OT

</dd>

</dl>

</div>

</dd></dl>

ot.mapping.joint_OT_mapping_linear(xs, xt, mu=1, eta=0.001, bias=False, verbose=False, verbose2=False, numItermax=100, numInnerItermax=10, stopInnerThr=1e-06, stopThr=1e-05, log=False, **kwargs)<a class="reference internal" href="../_modules/ot/mapping.html#joint_OT_mapping_linear">[source]</a><a class="headerlink" href="#ot.mapping.joint_OT_mapping_linear" title="Link to this definition"></a></dt>

<dd>Joint OT and linear mapping estimation as proposed in

[8].

The function solves the following optimization problem:

\[ \begin{align}\begin{aligned}\min_{\gamma,L}\quad \|L(\mathbf{X_s}) - n_s\gamma \mathbf{X_t} \|^2_F +

\mu \langle \gamma, \mathbf{M} \rangle_F + \eta \|L - \mathbf{I}\|^2_F\\s.t. \ \gamma \mathbf{1} = \mathbf{a}\\ \gamma^T \mathbf{1} = \mathbf{b}\\ \gamma \geq 0\end{aligned}\end{align} \]</div>

where :

<li>\(\mathbf{M}\) is the (<cite>ns</cite>, <cite>nt</cite>) squared euclidean cost matrix between samples in

<li>\(L\) is a \(d\times d\) linear operator that approximates the barycentric

mapping</li>

<li>\(\mathbf{I}\) is the identity matrix (neutral linear mapping)</li>

<li>\(\mathbf{a}\) and \(\mathbf{b}\) are uniform source and target weights</li>

</ul>

The problem consist in solving jointly an optimal transport matrix

\(\gamma\) and a linear mapping that fits the barycentric mapping

\(n_s\gamma \mathbf{X_t}\).

One can also estimate a mapping with constant bias (see supplementary

material of [8]) using the bias optional argument.

The algorithm used for solving the problem is the block coordinate

descent that alternates between updates of \(\mathbf{G}\) (using conditional gradient)

and the update of \(\mathbf{L}\) using a classical least square solver.

<dt class="field-odd">Parameters:</dt>

<li>xs (array-like (ns,d)) – samples in the source domain</li>

<li>xt (array-like (nt,d)) – samples in the target domain</li>

</ul>

</dd>

<dt class="field-even">Returns:</dt>

<li>gamma ((ns, nt) array-like) – Optimal transportation matrix for the given parameters</li>

<li>L ((d, d) array-like) – Linear mapping matrix ((\(d+1\), <cite>d</cite>) if bias)</li>

<li>log (dict) – log dictionary return only if log==True in parameters</li>

</ul>

</dd>

</dl>

References

[8]

M. Perrot, N. Courty, R. Flamary, A. Habrard,

“Mapping estimation for discrete optimal transport”,

Neural Information Processing Systems (NIPS), 2016.

</aside>

See also

</dd>

</dl>

</div>

</dd></dl>

ot.mapping.nearest_brenier_potential_fit(X, V, X_classes=None, a=None, b=None, strongly_convex_constant=0.6, gradient_lipschitz_constant=1.4, its=100, log=False, init_method='barycentric', solver=None)<a class="reference internal" href="../_modules/ot/mapping.html#nearest_brenier_potential_fit">[source]</a><a class="headerlink" href="#ot.mapping.nearest_brenier_potential_fit" title="Link to this definition"></a></dt>

<dd>Computes optimal values and gradients at X for a strongly convex potential \(\varphi\) with Lipschitz gradients

on the partitions defined by <cite>X_classes</cite>, where \(\varphi\) is optimal such that

\(\nabla \varphi \#\mu \approx \nu\), given samples \(X = x_1, \cdots, x_n \sim \mu\) and

\(V = v_1, \cdots, v_n \sim \nu\). Finding such a potential that has the desired regularity on the

partition \((E_k)_{k \in [K]}\) (given by the classes <cite>X_classes</cite>) is equivalent to finding optimal values

<cite>phi</cite> for the \(\varphi(x_i)\) and its gradients \(\nabla \varphi(x_i)\) (variable`G`).

In practice, these optimal values are found by solving the following problem

\begin{gather*}

\text{min} \sum_{i,j}\pi_{i,j}\|g_i - v_j\|_2^2 \\

g_1,\cdots, g_n \in \mathbb{R}^d,\; \varphi_1, \cdots, \varphi_n \in \mathbb{R},\; \pi \in \Pi(a, b) \\

\text{s.t.}\ \forall k \in [K],\; \forall i,j \in I_k: \\

\varphi_i-\varphi_j-\langle g_j, x_i-x_j\rangle \geq c_1\|g_i - g_j\|_2^2

+ c_2\|x_i-x_j\|_2^2 - c_3\langle g_j-g_i, x_j -x_i \rangle.

\end{gather*}</div>The constants \(c_1, c_2, c_3\) only depend on <cite>strongly_convex_constant</cite> and <cite>gradient_lipschitz_constant</cite>.

The constraint \(\pi \in \Pi(a, b)\) denotes the fact that the matrix \(\pi\) belong to the OT polytope

of marginals a and b. \(I_k\) is the subset of \([n]\) of the i such that \(x_i\) is in the

partition (or class) \(E_k\), i.e. <cite>X_classes[i] == k</cite>.

This problem is solved by alternating over the variable \(\pi\) and the variables \(\varphi_i, g_i\).

For \(\pi\), the problem is the standard discrete OT problem, and for \(\varphi_i, g_i\), the

problem is a convex QCQP solved using <code class="code docutils literal notranslate">cvxpy</code> (ECOS solver).

Accepts any compatible backend, but will perform the QCQP optimisation on Numpy arrays, and convert back at the end.

Warning

This function requires the CVXPY library

</div>

Warning

Accepts any backend but will convert to Numpy then back to the backend.

</div>

<dt class="field-odd">Parameters:</dt>

<li>X (array-like (n, d)) – reference points used to compute the optimal values phi and G</li>

<li>V (array-like (n, d)) – values of the gradients at the reference points X</li>

<li>X_classes (array-like (n,), optional) – classes of the reference points, defaults to a single class</li>

<li>a (array-like (n,), optional) – weights for the reference points X, defaults to uniform</li>

<li>b (array-like (n,), optional) – weights for the target points V, defaults to uniform</li>

<li>strongly_convex_constant (<a class="reference external" href="https://docs.python.org/3/library/functions.html#float" title="(in Python v3.14)">float</a>, optional) – constant for the strong convexity of the input potential phi, defaults to 0.6</li>

<li>gradient_lipschitz_constant (<a class="reference external" href="https://docs.python.org/3/library/functions.html#float" title="(in Python v3.14)">float</a>, optional) – constant for the Lipschitz property of the input gradient G, defaults to 1.4</li>

<li>its (<a class="reference external" href="https://docs.python.org/3/library/functions.html#int" title="(in Python v3.14)">int</a>, optional) – number of iterations, defaults to 100</li>

<li>init_method (<a class="reference external" href="https://docs.python.org/3/library/stdtypes.html#str" title="(in Python v3.14)">str</a>, optional) – ‘target’ initialises G=V, ‘barycentric’ initialises at the image of X by

the barycentric projection</li>

<li>solver (<a class="reference external" href="https://docs.python.org/3/library/stdtypes.html#str" title="(in Python v3.14)">str</a>, optional) – The CVXPY solver to use</li>

</ul>

</dd>

<dt class="field-even">Returns:</dt>

<li>phi (array-like (n,)) – optimal values of the potential at the points X</li>

<li>G (array-like (n, d)) – optimal values of the gradients at the points X</li>

<li>log (dict, optional) – If input log is true, a dictionary containing the values of the variables at each iteration, as well

as solver information</li>

</ul>

</dd>

</dl>

References

[58]

François-Pierre Paty, Alexandre d’Aspremont, and Marco Cuturi. Regularity as regularization:

Smooth and strongly convex brenier potentials in optimal transport. In International Conference

on Artificial Intelligence and Statistics, pages 1222–1232. PMLR, 2020.

</aside>

See also

<dt><a class="reference internal" href="#id13" title="ot.mapping.nearest_brenier_potential_predict_bounds"><code class="xref py py-obj docutils literal notranslate">ot.mapping.nearest_brenier_potential_predict_bounds</code></a></dt><dd>Predicting SSNB images on new source data

</dd>

<dt><a class="reference internal" href="ot.da.html#id90" title="ot.da.NearestBrenierPotential"><code class="xref py py-obj docutils literal notranslate">ot.da.NearestBrenierPotential</code></a></dt><dd>BaseTransport wrapper for SSNB

</dd>

</dl>

</div>

</dd></dl>

<h3>Examples using <code class="docutils literal notranslate">ot.mapping.nearest_brenier_potential_fit</code><a class="headerlink" href="#examples-using-ot-mapping-nearest-brenier-potential-fit" title="Link to this heading"></a></h3>

<a class="reference internal" href="../auto_examples/others/plot_SSNB.html">Smooth and Strongly Convex Nearest Brenier Potentials</a>

<div class="sphx-glr-thumbnail-title">Smooth and Strongly Convex Nearest Brenier Potentials</div>

</div></div><div class="sphx-glr-clear"></div><dl class="py function">

ot.mapping.nearest_brenier_potential_predict_bounds(X, phi, G, Y, X_classes=None, Y_classes=None, strongly_convex_constant=0.6, gradient_lipschitz_constant=1.4, log=False, solver=None)<a class="reference internal" href="../_modules/ot/mapping.html#nearest_brenier_potential_predict_bounds">[source]</a><a class="headerlink" href="#ot.mapping.nearest_brenier_potential_predict_bounds" title="Link to this definition"></a></dt>

<dd>Compute the values of the lower and upper bounding potentials at the input points Y, using the potential optimal

values phi at X and their gradients G at X. The ‘lower’ potential corresponds to the method from [58],

Equation 2, while the bounding property and ‘upper’ potential come from [59], Theorem 3.14 (taking into

account the fact that this theorem’s statement has a min instead of a max, which is a typo). Both potentials are

optimal for the SSNB problem.

If \(I_k\) is the subset of \([n]\) of the i such that \(x_i\) is in the partition (or class)

\(E_k\), for each \(y \in E_k\), this function solves the convex QCQP problems,

respectively for l: ‘lower’ and u: ‘upper’:

\begin{gather*}

(\varphi_{l}(x), \nabla \varphi_l(x)) = \text{argmin}\ t, \\

t\in \mathbb{R},\; g\in \mathbb{R}^d, \\

\text{s.t.} \forall j \in I_k,\; t-\varphi_j - \langle g_j, y-x_j \rangle \geq c_1\|g - g_j\|_2^2

+ c_2\|y-x_j\|_2^2 - c_3\langle g_j-g, x_j -y \rangle.

\end{gather*}</div><div class="math notranslate nohighlight">

\begin{gather*}

(\varphi_{u}(x), \nabla \varphi_u(x)) = \text{argmax}\ t, \\

t\in \mathbb{R},\; g\in \mathbb{R}^d, \\

\text{s.t.} \forall i \in I_k,\; \varphi_i^* -t - \langle g, x_i-y \rangle \geq c_1\|g_i - g\|_2^2

+ c_2\|x_i-y\|_2^2 - c_3\langle g-g_i, y -x_i \rangle.

Warning

This function requires the CVXPY library

</div>

Warning

Accepts any backend but will convert to Numpy then back to the backend.

</div>

<dt class="field-odd">Parameters:</dt>

<li>X (array-like (n, d)) – reference points used to compute the optimal values phi and G</li>

<li>X_classes (array-like (n,), optional) – classes of the reference points</li>

<li>phi (array-like (n,)) – optimal values of the potential at the points X</li>

<li>G (array-like (n, d)) – optimal values of the gradients at the points X</li>

<li>Y (array-like (m, d)) – input points</li>

<li>X_classes – classes of the reference points, defaults to a single class</li>

<li>Y_classes (array_like (m,), optional) – classes of the input points, defaults to a single class</li>

</ul>

</dd>

<dt class="field-even">Returns:</dt>

<li>phi_lu (array-like (2, m)) – values of the lower and upper bounding potentials at Y</li>

<li>G_lu (array-like (2, m, d)) – gradients of the lower and upper bounding potentials at Y</li>

<li>log (dict, optional) – If input log is true, a dictionary containing solver information</li>

</ul>

</dd>

</dl>

References

[58]

François-Pierre Paty, Alexandre d’Aspremont, and Marco Cuturi. Regularity as regularization:

Smooth and strongly convex brenier potentials in optimal transport. In International Conference

on Artificial Intelligence and Statistics, pages 1222–1232. PMLR, 2020.

</aside>

[59]

Adrien B Taylor. Convex interpolation and performance estimation of first-order methods for

convex optimization. PhD thesis, Catholic University of Louvain, Louvain-la-Neuve, Belgium,

2017.

</aside>

See also

<dt><a class="reference internal" href="#id11" title="ot.mapping.nearest_brenier_potential_fit"><code class="xref py py-obj docutils literal notranslate">ot.mapping.nearest_brenier_potential_fit</code></a></dt><dd>Fitting the SSNB on source and target data

</dd>

</dl>

</div>

</dd></dl>

</section>

<h3>Examples using <code class="docutils literal notranslate">ot.mapping.nearest_brenier_potential_predict_bounds</code><a class="headerlink" href="#examples-using-ot-mapping-nearest-brenier-potential-predict-bounds" title="Link to this heading"></a></h3>

<a class="reference internal" href="../auto_examples/others/plot_SSNB.html">Smooth and Strongly Convex Nearest Brenier Potentials</a>

<div class="sphx-glr-thumbnail-title">Smooth and Strongly Convex Nearest Brenier Potentials</div>

</div></div><div class="sphx-glr-clear"></div></section>

</section>

<dd>Joint OT and nonlinear mapping estimation with kernels as proposed in

[8].

The function solves the following optimization problem:

\[ \begin{align}\begin{aligned}\min_{\gamma, L\in\mathcal{H}}\quad \|L(\mathbf{X_s}) -

n_s\gamma \mathbf{X_t}\|^2_F + \mu \langle \gamma, \mathbf{M} \rangle_F +

\eta \|L\|^2_\mathcal{H}\\s.t. \ \gamma \mathbf{1} = \mathbf{a}\\ \gamma^T \mathbf{1} = \mathbf{b}\\ \gamma \geq 0\end{aligned}\end{align} \]</div>

where :

<li>\(\mathbf{M}\) is the (<cite>ns</cite>, <cite>nt</cite>) squared euclidean cost matrix between samples in

<li>\(L\) is a \(n_s \times d\) linear operator on a kernel matrix that

approximates the barycentric mapping</li>

<li>\(\mathbf{a}\) and \(\mathbf{b}\) are uniform source and target weights</li>

</ul>

The problem consist in solving jointly an optimal transport matrix

\(\gamma\) and the nonlinear mapping that fits the barycentric mapping

\(n_s\gamma \mathbf{X_t}\).

One can also estimate a mapping with constant bias (see supplementary

material of [8]) using the bias optional argument.

The algorithm used for solving the problem is the block coordinate

descent that alternates between updates of \(\mathbf{G}\) (using conditional gradient)

and the update of \(\mathbf{L}\) using a classical kernel least square solver.

<dt class="field-odd">Parameters:</dt>

<li>xs (array-like (ns,d)) – samples in the source domain</li>

<li>xt (array-like (nt,d)) – samples in the target domain</li>

</ul>

</dd>

<dt class="field-even">Returns:</dt>

<li>gamma ((ns, nt) array-like) – Optimal transportation matrix for the given parameters</li>

<li>L ((ns, d) array-like) – Nonlinear mapping matrix ((\(n_s+1\), <cite>d</cite>) if bias)</li>

<li>log (dict) – log dictionary return only if log==True in parameters</li>

</ul>

</dd>

</dl>

References

[8]

M. Perrot, N. Courty, R. Flamary, A. Habrard,

“Mapping estimation for discrete optimal transport”,

Neural Information Processing Systems (NIPS), 2016.

</aside>

See also

</dd>

</dl>

</div>

</dd></dl>

<dd>Joint OT and linear mapping estimation as proposed in

[8].

The function solves the following optimization problem:

\[ \begin{align}\begin{aligned}\min_{\gamma,L}\quad \|L(\mathbf{X_s}) - n_s\gamma \mathbf{X_t} \|^2_F +

\mu \langle \gamma, \mathbf{M} \rangle_F + \eta \|L - \mathbf{I}\|^2_F\\s.t. \ \gamma \mathbf{1} = \mathbf{a}\\ \gamma^T \mathbf{1} = \mathbf{b}\\ \gamma \geq 0\end{aligned}\end{align} \]</div>

where :

<li>\(\mathbf{M}\) is the (<cite>ns</cite>, <cite>nt</cite>) squared euclidean cost matrix between samples in

<li>\(L\) is a \(d\times d\) linear operator that approximates the barycentric

mapping</li>

<li>\(\mathbf{I}\) is the identity matrix (neutral linear mapping)</li>

<li>\(\mathbf{a}\) and \(\mathbf{b}\) are uniform source and target weights</li>

</ul>

The problem consist in solving jointly an optimal transport matrix

\(\gamma\) and a linear mapping that fits the barycentric mapping

\(n_s\gamma \mathbf{X_t}\).

One can also estimate a mapping with constant bias (see supplementary

material of [8]) using the bias optional argument.

The algorithm used for solving the problem is the block coordinate

descent that alternates between updates of \(\mathbf{G}\) (using conditional gradient)

and the update of \(\mathbf{L}\) using a classical least square solver.

<dt class="field-odd">Parameters:</dt>

<li>xs (array-like (ns,d)) – samples in the source domain</li>

<li>xt (array-like (nt,d)) – samples in the target domain</li>

</ul>

</dd>

<dt class="field-even">Returns:</dt>

<li>gamma ((ns, nt) array-like) – Optimal transportation matrix for the given parameters</li>

<li>L ((d, d) array-like) – Linear mapping matrix ((\(d+1\), <cite>d</cite>) if bias)</li>

<li>log (dict) – log dictionary return only if log==True in parameters</li>

</ul>

</dd>

</dl>

References

[8]

M. Perrot, N. Courty, R. Flamary, A. Habrard,

“Mapping estimation for discrete optimal transport”,

Neural Information Processing Systems (NIPS), 2016.

</aside>

See also

</dd>

</dl>

</div>

</dd></dl>

<dd>Computes optimal values and gradients at X for a strongly convex potential \(\varphi\) with Lipschitz gradients

on the partitions defined by <cite>X_classes</cite>, where \(\varphi\) is optimal such that

\(\nabla \varphi \#\mu \approx \nu\), given samples \(X = x_1, \cdots, x_n \sim \mu\) and

\(V = v_1, \cdots, v_n \sim \nu\). Finding such a potential that has the desired regularity on the

partition \((E_k)_{k \in [K]}\) (given by the classes <cite>X_classes</cite>) is equivalent to finding optimal values

<cite>phi</cite> for the \(\varphi(x_i)\) and its gradients \(\nabla \varphi(x_i)\) (variable`G`).

In practice, these optimal values are found by solving the following problem

\begin{gather*}

\text{min} \sum_{i,j}\pi_{i,j}\|g_i - v_j\|_2^2 \\

g_1,\cdots, g_n \in \mathbb{R}^d,\; \varphi_1, \cdots, \varphi_n \in \mathbb{R},\; \pi \in \Pi(a, b) \\

\text{s.t.}\ \forall k \in [K],\; \forall i,j \in I_k: \\

\varphi_i-\varphi_j-\langle g_j, x_i-x_j\rangle \geq c_1\|g_i - g_j\|_2^2

+ c_2\|x_i-x_j\|_2^2 - c_3\langle g_j-g_i, x_j -x_i \rangle.

partition (or class) \(E_k\), i.e. <cite>X_classes[i] == k</cite>.

For \(\pi\), the problem is the standard discrete OT problem, and for \(\varphi_i, g_i\), the

problem is a convex QCQP solved using <code class="code docutils literal notranslate">cvxpy</code> (ECOS solver).

Accepts any compatible backend, but will perform the QCQP optimisation on Numpy arrays, and convert back at the end.

Warning

This function requires the CVXPY library

</div>

Warning

Accepts any backend but will convert to Numpy then back to the backend.

</div>

<dt class="field-odd">Parameters:</dt>

<li>X (array-like (n, d)) – reference points used to compute the optimal values phi and G</li>

<li>V (array-like (n, d)) – values of the gradients at the reference points X</li>

<li>X_classes (array-like (n,), optional) – classes of the reference points, defaults to a single class</li>

<li>a (array-like (n,), optional) – weights for the reference points X, defaults to uniform</li>

<li>b (array-like (n,), optional) – weights for the target points V, defaults to uniform</li>

the barycentric projection</li>

</ul>

</dd>

<dt class="field-even">Returns:</dt>

<li>phi (array-like (n,)) – optimal values of the potential at the points X</li>

<li>G (array-like (n, d)) – optimal values of the gradients at the points X</li>

<li>log (dict, optional) – If input log is true, a dictionary containing the values of the variables at each iteration, as well

as solver information</li>

</ul>

</dd>

</dl>

References

[58]

François-Pierre Paty, Alexandre d’Aspremont, and Marco Cuturi. Regularity as regularization:

Smooth and strongly convex brenier potentials in optimal transport. In International Conference

on Artificial Intelligence and Statistics, pages 1222–1232. PMLR, 2020.

</aside>

See also

</dd>

</dl>

</div>

</dd></dl>

ot.mapping.nearest_brenier_potential_predict_bounds(X, phi, G, Y, X_classes=None, Y_classes=None, strongly_convex_constant=0.6, gradient_lipschitz_constant=1.4, log=False, solver=None)<a class="reference internal" href="../_modules/ot/mapping.html#nearest_brenier_potential_predict_bounds">[source]</a><a class="headerlink" href="#id13" title="Link to this definition"></a></dt>

<dd>Compute the values of the lower and upper bounding potentials at the input points Y, using the potential optimal

values phi at X and their gradients G at X. The ‘lower’ potential corresponds to the method from [58],

Equation 2, while the bounding property and ‘upper’ potential come from [59], Theorem 3.14 (taking into

account the fact that this theorem’s statement has a min instead of a max, which is a typo). Both potentials are

optimal for the SSNB problem.

\(E_k\), for each \(y \in E_k\), this function solves the convex QCQP problems,

respectively for l: ‘lower’ and u: ‘upper’:

\begin{gather*}

(\varphi_{l}(x), \nabla \varphi_l(x)) = \text{argmin}\ t, \\

t\in \mathbb{R},\; g\in \mathbb{R}^d, \\

\text{s.t.} \forall j \in I_k,\; t-\varphi_j - \langle g_j, y-x_j \rangle \geq c_1\|g - g_j\|_2^2

+ c_2\|y-x_j\|_2^2 - c_3\langle g_j-g, x_j -y \rangle.

\end{gather*}</div><div class="math notranslate nohighlight">

\begin{gather*}

(\varphi_{u}(x), \nabla \varphi_u(x)) = \text{argmax}\ t, \\

t\in \mathbb{R},\; g\in \mathbb{R}^d, \\

\text{s.t.} \forall i \in I_k,\; \varphi_i^* -t - \langle g, x_i-y \rangle \geq c_1\|g_i - g\|_2^2

+ c_2\|x_i-y\|_2^2 - c_3\langle g-g_i, y -x_i \rangle.

Warning

This function requires the CVXPY library

</div>

Warning

Accepts any backend but will convert to Numpy then back to the backend.

</div>

<dt class="field-odd">Parameters:</dt>

<li>X (array-like (n, d)) – reference points used to compute the optimal values phi and G</li>

<li>X_classes (array-like (n,), optional) – classes of the reference points</li>

<li>phi (array-like (n,)) – optimal values of the potential at the points X</li>

<li>G (array-like (n, d)) – optimal values of the gradients at the points X</li>

<li>Y (array-like (m, d)) – input points</li>

<li>X_classes – classes of the reference points, defaults to a single class</li>

<li>Y_classes (array_like (m,), optional) – classes of the input points, defaults to a single class</li>

</ul>

</dd>

<dt class="field-even">Returns:</dt>

<li>phi_lu (array-like (2, m)) – values of the lower and upper bounding potentials at Y</li>

<li>G_lu (array-like (2, m, d)) – gradients of the lower and upper bounding potentials at Y</li>

<li>log (dict, optional) – If input log is true, a dictionary containing solver information</li>

</ul>

</dd>

</dl>

References

[58]

François-Pierre Paty, Alexandre d’Aspremont, and Marco Cuturi. Regularity as regularization:

Smooth and strongly convex brenier potentials in optimal transport. In International Conference

on Artificial Intelligence and Statistics, pages 1222–1232. PMLR, 2020.

</aside>

[59]

Adrien B Taylor. Convex interpolation and performance estimation of first-order methods for

convex optimization. PhD thesis, Catholic University of Louvain, Louvain-la-Neuve, Belgium,

2017.

</aside>

See also

</dd>

</dl>

</div>

</dd></dl>

</section>

</div>

<a href="ot.lp.html" class="btn btn-neutral float-left" title="ot.lp" accesskey="p" rel="prev"> Previous</a>

</div>

<hr/>

</div>

Built with <a href="https://www.sphinx-doc.org/">Sphinx</a> using a

<a href="https://github.com/readthedocs/sphinx_rtd_theme">theme</a>

provided by <a href="https://readthedocs.org">Read the Docs</a>.

</footer>

</div>

</section>

</div>

<div class="rst-versions" data-toggle="rst-versions" role="note"

aria-label="versions">

<!-- add shift_up to the class for force viewing ,

data-toggle="rst-current-version" -->

Python Optimal Transport 0.9.7

<!-- versions

-->

<span class="rst-current-version" style="display: inline-block;padding:

0px;color:#fcfcfcab;float:left;font-size: 100%;">

Versions:

<a href="https://pythonot.github.io/"

style="padding: 3px;color:#fcfcfc;font-size: 100%;">Release</a>

<a href="https://pythonot.github.io/master"

style="padding: 3px;color:#fcfcfc;font-size: 100%;">Development</a>

<a href="https://github.com/PythonOT/POT"

style="padding: 3px;color:#fcfcfc;font-size: 100%;">Code</a>

<!--

<dl>

<dt>Versions</dt>

<dd><a href="https://pythonot.github.io/">Release</a></dd>

<dd><a href="https://pythonot.github.io/master">Development</a></dd>

<dt><a href="https://github.com/PythonOT/POT">Code on Github</a></dt>

</dl>

<hr>

</div>

</div>-->

</div><script>

jQuery(function () {

SphinxRtdTheme.Navigation.enable(true);

});

</script>

</body>

</html>

FilesExpand file tree

ot.mapping.html

Latest commit

History

ot.mapping.html

File metadata and controls