Hide keyboard shortcuts

Hot-keys on this page

r m x p   toggle line displays

j k   next/prev highlighted chunk

0   (zero) top of page

1   (one) first highlighted chunk

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

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

855

856

857

858

859

860

861

862

863

864

865

866

867

868

869

870

871

872

873

874

875

876

877

878

879

880

881

882

883

884

885

886

887

888

889

890

891

892

893

894

895

896

897

898

899

900

901

902

903

904

905

906

907

908

909

910

911

912

913

914

915

916

917

918

919

920

921

922

923

924

925

926

927

928

929

930

931

932

933

934

935

936

937

938

939

940

941

942

943

944

945

946

947

948

949

950

951

952

953

954

955

956

957

958

959

960

961

962

963

964

965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987

988

989

990

991

992

993

994

995

996

997

998

999

1000

1001

1002

1003

1004

1005

1006

1007

1008

1009

1010

1011

1012

1013

1014

1015

1016

1017

1018

1019

1020

1021

1022

1023

1024

1025

1026

1027

1028

1029

1030

1031

1032

1033

1034

1035

1036

1037

1038

1039

1040

1041

1042

1043

1044

1045

1046

1047

1048

1049

1050

1051

1052

1053

1054

1055

1056

1057

1058

1059

1060

1061

1062

1063

1064

1065

1066

1067

1068

1069

1070

1071

1072

1073

1074

1075

1076

1077

1078

1079

1080

1081

1082

1083

1084

1085

1086

1087

1088

1089

1090

1091

1092

1093

1094

1095

1096

1097

1098

1099

1100

1101

1102

1103

1104

1105

1106

1107

1108

1109

1110

1111

1112

1113

1114

1115

1116

1117

1118

1119

1120

1121

1122

1123

1124

1125

1126

1127

1128

1129

1130

1131

1132

1133

1134

1135

1136

1137

1138

1139

1140

1141

1142

1143

1144

1145

1146

1147

1148

1149

1150

1151

1152

1153

1154

1155

1156

1157

1158

1159

1160

1161

1162

1163

1164

1165

1166

1167

1168

1169

1170

1171

1172

1173

1174

1175

1176

1177

1178

1179

1180

1181

1182

1183

1184

1185

1186

1187

1188

1189

1190

1191

1192

1193

1194

1195

1196

1197

1198

1199

1200

1201

1202

1203

1204

1205

1206

1207

1208

1209

1210

1211

1212

1213

1214

1215

1216

1217

1218

1219

1220

1221

1222

1223

1224

1225

1226

1227

1228

1229

1230

1231

1232

1233

1234

1235

1236

1237

1238

1239

1240

1241

1242

1243

1244

1245

1246

1247

1248

1249

1250

1251

1252

1253

1254

1255

1256

1257

1258

1259

1260

1261

1262

1263

1264

1265

1266

1267

1268

1269

1270

1271

1272

1273

1274

1275

1276

1277

1278

1279

1280

1281

1282

1283

1284

1285

1286

1287

1288

1289

1290

1291

1292

1293

1294

1295

1296

1297

1298

1299

1300

1301

1302

1303

1304

1305

1306

1307

1308

1309

1310

1311

1312

1313

1314

1315

1316

1317

1318

1319

1320

1321

1322

1323

1324

1325

1326

1327

1328

1329

1330

1331

1332

1333

1334

1335

1336

1337

1338

1339

1340

1341

1342

1343

1344

1345

1346

1347

1348

1349

1350

1351

1352

1353

1354

1355

1356

1357

1358

1359

1360

1361

1362

1363

1364

1365

1366

1367

1368

1369

1370

1371

1372

1373

1374

1375

1376

1377

1378

1379

1380

1381

1382

1383

1384

1385

1386

1387

1388

1389

1390

1391

1392

1393

1394

1395

1396

1397

1398

1399

1400

1401

1402

1403

1404

1405

1406

1407

1408

1409

1410

1411

1412

1413

1414

1415

1416

1417

1418

1419

1420

1421

1422

1423

1424

1425

1426

1427

1428

1429

1430

1431

1432

1433

1434

1435

1436

1437

1438

1439

1440

1441

1442

1443

1444

1445

1446

1447

1448

1449

1450

1451

1452

1453

1454

1455

1456

1457

1458

1459

1460

1461

1462

1463

1464

1465

1466

1467

1468

1469

1470

1471

1472

1473

1474

1475

1476

1477

1478

1479

1480

1481

1482

1483

1484

1485

1486

1487

1488

1489

1490

1491

1492

1493

1494

1495

1496

1497

1498

1499

1500

1501

1502

1503

1504

1505

1506

1507

1508

1509

1510

1511

1512

1513

1514

1515

1516

1517

1518

1519

1520

1521

1522

1523

1524

1525

1526

1527

1528

1529

1530

1531

1532

1533

1534

1535

1536

1537

1538

1539

1540

1541

1542

1543

1544

1545

1546

1547

1548

1549

1550

1551

1552

1553

1554

1555

1556

1557

1558

1559

1560

1561

1562

1563

1564

1565

1566

1567

1568

1569

1570

1571

1572

1573

1574

1575

1576

1577

1578

1579

1580

1581

1582

1583

1584

1585

1586

1587

1588

1589

1590

1591

1592

1593

1594

1595

1596

1597

1598

1599

1600

1601

1602

1603

1604

1605

1606

1607

1608

1609

1610

1611

1612

1613

1614

1615

1616

1617

1618

1619

1620

1621

1622

1623

1624

1625

1626

1627

1628

1629

1630

1631

1632

1633

1634

1635

1636

1637

1638

1639

1640

1641

1642

1643

1644

1645

1646

1647

1648

1649

1650

1651

1652

1653

1654

1655

1656

1657

1658

1659

1660

1661

1662

1663

1664

1665

1666

1667

1668

1669

1670

1671

1672

1673

1674

1675

1676

1677

1678

1679

1680

1681

1682

1683

1684

1685

1686

1687

1688

1689

1690

1691

1692

1693

1694

1695

1696

1697

1698

1699

1700

1701

1702

1703

1704

1705

1706

1707

1708

1709

1710

1711

1712

1713

1714

1715

1716

1717

1718

1719

1720

1721

1722

1723

1724

1725

1726

1727

1728

1729

1730

1731

1732

1733

1734

1735

1736

1737

1738

1739

1740

1741

1742

1743

1744

1745

1746

1747

1748

1749

1750

1751

1752

1753

1754

1755

1756

1757

1758

1759

1760

1761

1762

1763

1764

1765

1766

1767

1768

1769

1770

1771

1772

1773

1774

1775

1776

1777

1778

1779

1780

1781

1782

1783

1784

1785

1786

1787

1788

1789

1790

1791

1792

1793

1794

1795

1796

1797

1798

1799

1800

1801

1802

1803

1804

1805

1806

1807

1808

1809

1810

1811

1812

1813

1814

1815

1816

1817

1818

1819

1820

1821

1822

1823

1824

1825

1826

1827

1828

1829

1830

1831

1832

1833

1834

1835

1836

1837

1838

1839

1840

1841

1842

1843

1844

1845

1846

1847

1848

1849

1850

1851

1852

1853

1854

1855

1856

1857

1858

1859

1860

1861

1862

1863

1864

1865

1866

1867

1868

1869

1870

1871

1872

1873

1874

1875

1876

1877

1878

1879

1880

1881

1882

1883

1884

1885

1886

1887

1888

1889

1890

1891

1892

1893

1894

1895

1896

1897

1898

1899

1900

1901

1902

1903

1904

1905

1906

1907

1908

1909

1910

1911

1912

1913

1914

1915

1916

1917

1918

1919

1920

1921

1922

1923

1924

1925

1926

1927

1928

1929

1930

1931

1932

1933

1934

1935

1936

1937

1938

1939

1940

1941

1942

1943

1944

1945

1946

1947

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

1958

1959

1960

1961

1962

1963

1964

1965

1966

1967

1968

1969

1970

1971

1972

1973

1974

1975

1976

1977

1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

2031

2032

2033

2034

2035

2036

2037

2038

2039

2040

2041

2042

2043

2044

2045

2046

2047

2048

2049

2050

2051

2052

2053

2054

2055

2056

2057

2058

2059

2060

2061

2062

2063

2064

2065

2066

2067

2068

2069

2070

2071

2072

2073

2074

2075

2076

2077

2078

2079

2080

2081

2082

2083

2084

2085

2086

2087

2088

2089

2090

2091

2092

2093

2094

2095

2096

2097

2098

2099

2100

2101

2102

2103

2104

2105

2106

2107

2108

2109

2110

2111

2112

2113

2114

2115

2116

2117

2118

2119

2120

2121

2122

2123

2124

2125

2126

2127

2128

2129

2130

2131

2132

2133

2134

2135

2136

2137

2138

2139

2140

2141

2142

2143

2144

2145

2146

2147

2148

2149

2150

2151

2152

2153

2154

2155

2156

2157

2158

2159

2160

2161

2162

2163

2164

2165

2166

2167

2168

2169

2170

2171

2172

2173

2174

2175

2176

2177

2178

2179

2180

2181

2182

2183

2184

2185

2186

2187

2188

2189

2190

2191

2192

2193

2194

2195

2196

2197

2198

2199

2200

2201

2202

2203

2204

2205

2206

2207

2208

2209

2210

2211

2212

2213

2214

2215

2216

2217

2218

2219

2220

2221

2222

2223

2224

2225

2226

2227

2228

2229

2230

2231

2232

2233

2234

2235

2236

2237

2238

2239

2240

2241

2242

2243

2244

2245

2246

2247

2248

2249

2250

2251

2252

2253

2254

2255

2256

2257

2258

2259

2260

2261

2262

2263

2264

2265

2266

2267

2268

2269

2270

2271

2272

2273

2274

2275

2276

2277

2278

2279

2280

2281

2282

2283

2284

2285

2286

2287

2288

2289

2290

2291

2292

2293

2294

2295

2296

2297

2298

2299

2300

2301

2302

2303

2304

2305

2306

2307

2308

2309

2310

2311

2312

2313

2314

2315

2316

2317

2318

2319

2320

2321

2322

2323

2324

2325

2326

2327

2328

2329

2330

2331

2332

2333

2334

2335

2336

2337

2338

2339

2340

2341

2342

2343

2344

2345

2346

2347

2348

2349

2350

2351

2352

2353

2354

2355

2356

2357

2358

2359

2360

2361

2362

2363

2364

2365

2366

2367

2368

2369

2370

2371

2372

2373

2374

2375

2376

2377

2378

2379

2380

2381

2382

2383

2384

2385

2386

2387

2388

2389

2390

2391

2392

2393

2394

2395

2396

2397

2398

2399

2400

2401

2402

2403

2404

2405

2406

2407

2408

2409

2410

2411

2412

2413

2414

2415

2416

2417

2418

2419

2420

2421

2422

2423

2424

2425

2426

2427

2428

2429

2430

2431

2432

2433

2434

2435

2436

2437

2438

2439

2440

2441

2442

2443

2444

2445

2446

2447

2448

2449

2450

2451

2452

2453

2454

2455

2456

2457

2458

2459

2460

2461

2462

2463

2464

2465

2466

2467

2468

2469

2470

2471

2472

2473

2474

2475

2476

2477

2478

2479

2480

2481

2482

2483

2484

2485

2486

2487

2488

2489

2490

2491

2492

2493

2494

2495

2496

2497

2498

2499

2500

2501

2502

2503

2504

2505

2506

2507

2508

2509

2510

2511

2512

2513

2514

2515

2516

2517

2518

2519

2520

2521

2522

2523

2524

2525

2526

2527

2528

2529

2530

2531

2532

2533

2534

2535

2536

2537

2538

2539

2540

2541

2542

2543

2544

2545

2546

2547

2548

2549

2550

2551

2552

2553

2554

2555

2556

2557

2558

2559

2560

2561

2562

2563

2564

2565

2566

2567

2568

2569

2570

2571

2572

2573

2574

2575

2576

2577

2578

2579

2580

2581

2582

2583

2584

2585

2586

2587

2588

2589

2590

2591

2592

2593

2594

2595

2596

2597

2598

2599

2600

2601

2602

2603

2604

2605

2606

2607

2608

2609

2610

2611

2612

2613

2614

2615

2616

2617

2618

2619

2620

2621

2622

2623

2624

2625

2626

2627

2628

2629

2630

2631

2632

2633

2634

2635

2636

2637

2638

2639

2640

2641

2642

2643

2644

2645

2646

2647

2648

2649

2650

2651

2652

2653

2654

2655

2656

2657

2658

2659

2660

2661

2662

2663

2664

2665

2666

2667

2668

2669

2670

2671

2672

2673

2674

2675

2676

2677

2678

2679

2680

2681

2682

2683

2684

2685

2686

2687

2688

2689

2690

2691

2692

2693

2694

2695

2696

2697

2698

2699

2700

2701

2702

2703

2704

2705

2706

2707

2708

2709

2710

2711

2712

2713

2714

2715

2716

2717

2718

2719

2720

2721

2722

2723

2724

2725

2726

2727

2728

2729

2730

2731

2732

2733

2734

2735

2736

2737

2738

2739

2740

2741

2742

2743

2744

2745

2746

2747

2748

2749

2750

2751

2752

2753

2754

2755

2756

2757

2758

2759

2760

2761

2762

2763

2764

2765

2766

2767

2768

2769

2770

2771

2772

2773

2774

2775

2776

2777

2778

2779

2780

2781

2782

2783

2784

2785

2786

2787

2788

2789

2790

2791

2792

2793

2794

2795

2796

2797

2798

2799

2800

2801

2802

2803

2804

2805

2806

2807

2808

2809

2810

2811

2812

2813

2814

2815

2816

2817

2818

2819

2820

2821

2822

2823

2824

2825

2826

2827

2828

2829

2830

2831

2832

2833

2834

""" 

A collection of utility functions and classes. Originally, many 

(but not all) were from the Python Cookbook -- hence the name cbook. 

 

This module is safe to import from anywhere within matplotlib; 

it imports matplotlib only at runtime. 

""" 

 

from __future__ import absolute_import, division, print_function 

 

import six 

from six.moves import xrange, zip 

import collections 

import contextlib 

import datetime 

import errno 

import functools 

import glob 

import gzip 

import io 

from itertools import repeat 

import locale 

import numbers 

import operator 

import os 

import re 

import sys 

import time 

import traceback 

import types 

import warnings 

from weakref import ref, WeakKeyDictionary 

 

import numpy as np 

 

import matplotlib 

from .deprecation import deprecated, warn_deprecated 

from .deprecation import mplDeprecation, MatplotlibDeprecationWarning 

 

 

def unicode_safe(s): 

 

if isinstance(s, bytes): 

try: 

# On some systems, locale.getpreferredencoding returns None, 

# which can break unicode; and the sage project reports that 

# some systems have incorrect locale specifications, e.g., 

# an encoding instead of a valid locale name. Another 

# pathological case that has been reported is an empty string. 

# On some systems, getpreferredencoding sets the locale, which has 

# side effects. Passing False eliminates those side effects. 

preferredencoding = locale.getpreferredencoding( 

matplotlib.rcParams['axes.formatter.use_locale']).strip() 

if not preferredencoding: 

preferredencoding = None 

except (ValueError, ImportError, AttributeError): 

preferredencoding = None 

 

if preferredencoding is None: 

return six.text_type(s) 

else: 

return six.text_type(s, preferredencoding) 

return s 

 

 

@deprecated('2.1') 

class converter(object): 

""" 

Base class for handling string -> python type with support for 

missing values 

""" 

def __init__(self, missing='Null', missingval=None): 

self.missing = missing 

self.missingval = missingval 

 

def __call__(self, s): 

if s == self.missing: 

return self.missingval 

return s 

 

def is_missing(self, s): 

return not s.strip() or s == self.missing 

 

 

@deprecated('2.1') 

class tostr(converter): 

"""convert to string or None""" 

def __init__(self, missing='Null', missingval=''): 

converter.__init__(self, missing=missing, missingval=missingval) 

 

 

@deprecated('2.1') 

class todatetime(converter): 

"""convert to a datetime or None""" 

def __init__(self, fmt='%Y-%m-%d', missing='Null', missingval=None): 

'use a :func:`time.strptime` format string for conversion' 

converter.__init__(self, missing, missingval) 

self.fmt = fmt 

 

def __call__(self, s): 

if self.is_missing(s): 

return self.missingval 

tup = time.strptime(s, self.fmt) 

return datetime.datetime(*tup[:6]) 

 

 

@deprecated('2.1') 

class todate(converter): 

"""convert to a date or None""" 

def __init__(self, fmt='%Y-%m-%d', missing='Null', missingval=None): 

"""use a :func:`time.strptime` format string for conversion""" 

converter.__init__(self, missing, missingval) 

self.fmt = fmt 

 

def __call__(self, s): 

if self.is_missing(s): 

return self.missingval 

tup = time.strptime(s, self.fmt) 

return datetime.date(*tup[:3]) 

 

 

@deprecated('2.1') 

class tofloat(converter): 

"""convert to a float or None""" 

def __init__(self, missing='Null', missingval=None): 

converter.__init__(self, missing) 

self.missingval = missingval 

 

def __call__(self, s): 

if self.is_missing(s): 

return self.missingval 

return float(s) 

 

 

@deprecated('2.1') 

class toint(converter): 

"""convert to an int or None""" 

def __init__(self, missing='Null', missingval=None): 

converter.__init__(self, missing) 

 

def __call__(self, s): 

if self.is_missing(s): 

return self.missingval 

return int(s) 

 

 

class _BoundMethodProxy(object): 

""" 

Our own proxy object which enables weak references to bound and unbound 

methods and arbitrary callables. Pulls information about the function, 

class, and instance out of a bound method. Stores a weak reference to the 

instance to support garbage collection. 

 

@organization: IBM Corporation 

@copyright: Copyright (c) 2005, 2006 IBM Corporation 

@license: The BSD License 

 

Minor bugfixes by Michael Droettboom 

""" 

def __init__(self, cb): 

self._hash = hash(cb) 

self._destroy_callbacks = [] 

try: 

try: 

if six.PY3: 

self.inst = ref(cb.__self__, self._destroy) 

else: 

self.inst = ref(cb.im_self, self._destroy) 

except TypeError: 

self.inst = None 

if six.PY3: 

self.func = cb.__func__ 

self.klass = cb.__self__.__class__ 

else: 

self.func = cb.im_func 

self.klass = cb.im_class 

except AttributeError: 

self.inst = None 

self.func = cb 

self.klass = None 

 

def add_destroy_callback(self, callback): 

self._destroy_callbacks.append(_BoundMethodProxy(callback)) 

 

def _destroy(self, wk): 

for callback in self._destroy_callbacks: 

try: 

callback(self) 

except ReferenceError: 

pass 

 

def __getstate__(self): 

d = self.__dict__.copy() 

# de-weak reference inst 

inst = d['inst'] 

if inst is not None: 

d['inst'] = inst() 

return d 

 

def __setstate__(self, statedict): 

self.__dict__ = statedict 

inst = statedict['inst'] 

# turn inst back into a weakref 

if inst is not None: 

self.inst = ref(inst) 

 

def __call__(self, *args, **kwargs): 

""" 

Proxy for a call to the weak referenced object. Take 

arbitrary params to pass to the callable. 

 

Raises `ReferenceError`: When the weak reference refers to 

a dead object 

""" 

if self.inst is not None and self.inst() is None: 

raise ReferenceError 

elif self.inst is not None: 

# build a new instance method with a strong reference to the 

# instance 

 

mtd = types.MethodType(self.func, self.inst()) 

 

else: 

# not a bound method, just return the func 

mtd = self.func 

# invoke the callable and return the result 

return mtd(*args, **kwargs) 

 

def __eq__(self, other): 

""" 

Compare the held function and instance with that held by 

another proxy. 

""" 

try: 

if self.inst is None: 

return self.func == other.func and other.inst is None 

else: 

return self.func == other.func and self.inst() == other.inst() 

except Exception: 

return False 

 

def __ne__(self, other): 

""" 

Inverse of __eq__. 

""" 

return not self.__eq__(other) 

 

def __hash__(self): 

return self._hash 

 

 

def _exception_printer(exc): 

traceback.print_exc() 

 

 

class CallbackRegistry(object): 

"""Handle registering and disconnecting for a set of signals and callbacks: 

 

>>> def oneat(x): 

... print('eat', x) 

>>> def ondrink(x): 

... print('drink', x) 

 

>>> from matplotlib.cbook import CallbackRegistry 

>>> callbacks = CallbackRegistry() 

 

>>> id_eat = callbacks.connect('eat', oneat) 

>>> id_drink = callbacks.connect('drink', ondrink) 

 

>>> callbacks.process('drink', 123) 

drink 123 

>>> callbacks.process('eat', 456) 

eat 456 

>>> callbacks.process('be merry', 456) # nothing will be called 

>>> callbacks.disconnect(id_eat) 

>>> callbacks.process('eat', 456) # nothing will be called 

 

In practice, one should always disconnect all callbacks when they 

are no longer needed to avoid dangling references (and thus memory 

leaks). However, real code in matplotlib rarely does so, and due 

to its design, it is rather difficult to place this kind of code. 

To get around this, and prevent this class of memory leaks, we 

instead store weak references to bound methods only, so when the 

destination object needs to die, the CallbackRegistry won't keep 

it alive. The Python stdlib weakref module can not create weak 

references to bound methods directly, so we need to create a proxy 

object to handle weak references to bound methods (or regular free 

functions). This technique was shared by Peter Parente on his 

`"Mindtrove" blog 

<http://mindtrove.info/python-weak-references/>`_. 

 

 

Parameters 

---------- 

exception_handler : callable, optional 

If provided must have signature :: 

 

def handler(exc: Exception) -> None: 

 

If not None this function will be called with any `Exception` 

subclass raised by the callbacks in `CallbackRegistry.process`. 

The handler may either consume the exception or re-raise. 

 

The callable must be pickle-able. 

 

The default handler is :: 

 

def h(exc): 

traceback.print_exc() 

 

""" 

def __init__(self, exception_handler=_exception_printer): 

self.exception_handler = exception_handler 

self.callbacks = dict() 

self._cid = 0 

self._func_cid_map = {} 

 

# In general, callbacks may not be pickled; thus, we simply recreate an 

# empty dictionary at unpickling. In order to ensure that `__setstate__` 

# (which just defers to `__init__`) is called, `__getstate__` must 

# return a truthy value (for pickle protocol>=3, i.e. Py3, the 

# *actual* behavior is that `__setstate__` will be called as long as 

# `__getstate__` does not return `None`, but this is undocumented -- see 

# http://bugs.python.org/issue12290). 

 

def __getstate__(self): 

return {'exception_handler': self.exception_handler} 

 

def __setstate__(self, state): 

self.__init__(**state) 

 

def connect(self, s, func): 

"""Register *func* to be called when signal *s* is generated. 

""" 

self._func_cid_map.setdefault(s, WeakKeyDictionary()) 

# Note proxy not needed in python 3. 

# TODO rewrite this when support for python2.x gets dropped. 

proxy = _BoundMethodProxy(func) 

if proxy in self._func_cid_map[s]: 

return self._func_cid_map[s][proxy] 

 

proxy.add_destroy_callback(self._remove_proxy) 

self._cid += 1 

cid = self._cid 

self._func_cid_map[s][proxy] = cid 

self.callbacks.setdefault(s, dict()) 

self.callbacks[s][cid] = proxy 

return cid 

 

def _remove_proxy(self, proxy): 

for signal, proxies in list(six.iteritems(self._func_cid_map)): 

try: 

del self.callbacks[signal][proxies[proxy]] 

except KeyError: 

pass 

 

if len(self.callbacks[signal]) == 0: 

del self.callbacks[signal] 

del self._func_cid_map[signal] 

 

def disconnect(self, cid): 

"""Disconnect the callback registered with callback id *cid*. 

""" 

for eventname, callbackd in list(six.iteritems(self.callbacks)): 

try: 

del callbackd[cid] 

except KeyError: 

continue 

else: 

for signal, functions in list( 

six.iteritems(self._func_cid_map)): 

for function, value in list(six.iteritems(functions)): 

if value == cid: 

del functions[function] 

return 

 

def process(self, s, *args, **kwargs): 

""" 

Process signal *s*. 

 

All of the functions registered to receive callbacks on *s* will be 

called with ``*args`` and ``**kwargs``. 

""" 

if s in self.callbacks: 

for cid, proxy in list(six.iteritems(self.callbacks[s])): 

try: 

proxy(*args, **kwargs) 

except ReferenceError: 

self._remove_proxy(proxy) 

# this does not capture KeyboardInterrupt, SystemExit, 

# and GeneratorExit 

except Exception as exc: 

if self.exception_handler is not None: 

self.exception_handler(exc) 

else: 

raise 

 

 

class silent_list(list): 

""" 

override repr when returning a list of matplotlib artists to 

prevent long, meaningless output. This is meant to be used for a 

homogeneous list of a given type 

""" 

def __init__(self, type, seq=None): 

self.type = type 

if seq is not None: 

self.extend(seq) 

 

def __repr__(self): 

return '<a list of %d %s objects>' % (len(self), self.type) 

 

def __str__(self): 

return repr(self) 

 

def __getstate__(self): 

# store a dictionary of this SilentList's state 

return {'type': self.type, 'seq': self[:]} 

 

def __setstate__(self, state): 

self.type = state['type'] 

self.extend(state['seq']) 

 

 

class IgnoredKeywordWarning(UserWarning): 

""" 

A class for issuing warnings about keyword arguments that will be ignored 

by matplotlib 

""" 

pass 

 

 

def local_over_kwdict(local_var, kwargs, *keys): 

""" 

Enforces the priority of a local variable over potentially conflicting 

argument(s) from a kwargs dict. The following possible output values are 

considered in order of priority: 

 

local_var > kwargs[keys[0]] > ... > kwargs[keys[-1]] 

 

The first of these whose value is not None will be returned. If all are 

None then None will be returned. Each key in keys will be removed from the 

kwargs dict in place. 

 

Parameters 

---------- 

local_var: any object 

The local variable (highest priority) 

 

kwargs: dict 

Dictionary of keyword arguments; modified in place 

 

keys: str(s) 

Name(s) of keyword arguments to process, in descending order of 

priority 

 

Returns 

------- 

out: any object 

Either local_var or one of kwargs[key] for key in keys 

 

Raises 

------ 

IgnoredKeywordWarning 

For each key in keys that is removed from kwargs but not used as 

the output value 

 

""" 

out = local_var 

for key in keys: 

kwarg_val = kwargs.pop(key, None) 

if kwarg_val is not None: 

if out is None: 

out = kwarg_val 

else: 

warnings.warn('"%s" keyword argument will be ignored' % key, 

IgnoredKeywordWarning) 

return out 

 

 

def strip_math(s): 

"""remove latex formatting from mathtext""" 

remove = (r'\mathdefault', r'\rm', r'\cal', r'\tt', r'\it', '\\', '{', '}') 

s = s[1:-1] 

for r in remove: 

s = s.replace(r, '') 

return s 

 

 

class Bunch(object): 

""" 

Often we want to just collect a bunch of stuff together, naming each 

item of the bunch; a dictionary's OK for that, but a small do- nothing 

class is even handier, and prettier to use. Whenever you want to 

group a few variables:: 

 

>>> point = Bunch(datum=2, squared=4, coord=12) 

>>> point.datum 

 

By: Alex Martelli 

From: https://code.activestate.com/recipes/121294/ 

""" 

def __init__(self, **kwds): 

self.__dict__.update(kwds) 

 

def __repr__(self): 

return 'Bunch(%s)' % ', '.join( 

'%s=%s' % kv for kv in six.iteritems(vars(self))) 

 

 

@deprecated('2.1') 

def unique(x): 

"""Return a list of unique elements of *x*""" 

return list(set(x)) 

 

 

def iterable(obj): 

"""return true if *obj* is iterable""" 

try: 

iter(obj) 

except TypeError: 

return False 

return True 

 

 

@deprecated('2.1') 

def is_string_like(obj): 

"""Return True if *obj* looks like a string""" 

# (np.str_ == np.unicode_ on Py3). 

return isinstance(obj, (six.string_types, np.str_, np.unicode_)) 

 

 

@deprecated('2.1') 

def is_sequence_of_strings(obj): 

"""Returns true if *obj* is iterable and contains strings""" 

if not iterable(obj): 

return False 

if is_string_like(obj) and not isinstance(obj, np.ndarray): 

try: 

obj = obj.values 

except AttributeError: 

# not pandas 

return False 

for o in obj: 

if not is_string_like(o): 

return False 

return True 

 

 

def is_hashable(obj): 

"""Returns true if *obj* can be hashed""" 

try: 

hash(obj) 

except TypeError: 

return False 

return True 

 

 

def is_writable_file_like(obj): 

"""return true if *obj* looks like a file object with a *write* method""" 

return callable(getattr(obj, 'write', None)) 

 

 

def file_requires_unicode(x): 

""" 

Returns `True` if the given writable file-like object requires Unicode 

to be written to it. 

""" 

try: 

x.write(b'') 

except TypeError: 

return True 

else: 

return False 

 

 

@deprecated('2.1') 

def is_scalar(obj): 

"""return true if *obj* is not string like and is not iterable""" 

return not isinstance(obj, six.string_types) and not iterable(obj) 

 

 

def is_numlike(obj): 

"""return true if *obj* looks like a number""" 

return isinstance(obj, (numbers.Number, np.number)) 

 

 

def to_filehandle(fname, flag='rU', return_opened=False, encoding=None): 

""" 

*fname* can be an `os.PathLike` or a file handle. Support for gzipped 

files is automatic, if the filename ends in .gz. *flag* is a 

read/write flag for :func:`file` 

""" 

if hasattr(os, "PathLike") and isinstance(fname, os.PathLike): 

return to_filehandle( 

os.fspath(fname), 

flag=flag, return_opened=return_opened, encoding=encoding) 

if isinstance(fname, six.string_types): 

if fname.endswith('.gz'): 

# get rid of 'U' in flag for gzipped files. 

flag = flag.replace('U', '') 

fh = gzip.open(fname, flag) 

elif fname.endswith('.bz2'): 

# python may not be complied with bz2 support, 

# bury import until we need it 

import bz2 

# get rid of 'U' in flag for bz2 files 

flag = flag.replace('U', '') 

fh = bz2.BZ2File(fname, flag) 

else: 

fh = io.open(fname, flag, encoding=encoding) 

opened = True 

elif hasattr(fname, 'seek'): 

fh = fname 

opened = False 

else: 

raise ValueError('fname must be a PathLike or file handle') 

if return_opened: 

return fh, opened 

return fh 

 

 

@contextlib.contextmanager 

def open_file_cm(path_or_file, mode="r", encoding=None): 

r"""Pass through file objects and context-manage `.PathLike`\s.""" 

fh, opened = to_filehandle(path_or_file, mode, True, encoding) 

if opened: 

with fh: 

yield fh 

else: 

yield fh 

 

 

def is_scalar_or_string(val): 

"""Return whether the given object is a scalar or string like.""" 

return isinstance(val, six.string_types) or not iterable(val) 

 

 

def _string_to_bool(s): 

"""Parses the string argument as a boolean""" 

if not isinstance(s, six.string_types): 

return bool(s) 

warn_deprecated("2.2", "Passing one of 'on', 'true', 'off', 'false' as a " 

"boolean is deprecated; use an actual boolean " 

"(True/False) instead.") 

if s.lower() in ['on', 'true']: 

return True 

if s.lower() in ['off', 'false']: 

return False 

raise ValueError('String "%s" must be one of: ' 

'"on", "off", "true", or "false"' % s) 

 

 

def get_sample_data(fname, asfileobj=True): 

""" 

Return a sample data file. *fname* is a path relative to the 

`mpl-data/sample_data` directory. If *asfileobj* is `True` 

return a file object, otherwise just a file path. 

 

Set the rc parameter examples.directory to the directory where we should 

look, if sample_data files are stored in a location different than 

default (which is 'mpl-data/sample_data` at the same level of 'matplotlib` 

Python module files). 

 

If the filename ends in .gz, the file is implicitly ungzipped. 

""" 

if matplotlib.rcParams['examples.directory']: 

root = matplotlib.rcParams['examples.directory'] 

else: 

root = os.path.join(matplotlib._get_data_path(), 'sample_data') 

path = os.path.join(root, fname) 

 

if asfileobj: 

if (os.path.splitext(fname)[-1].lower() in 

('.csv', '.xrc', '.txt')): 

mode = 'r' 

else: 

mode = 'rb' 

 

base, ext = os.path.splitext(fname) 

if ext == '.gz': 

return gzip.open(path, mode) 

else: 

return open(path, mode) 

else: 

return path 

 

 

def flatten(seq, scalarp=is_scalar_or_string): 

""" 

Returns a generator of flattened nested containers 

 

For example: 

 

>>> from matplotlib.cbook import flatten 

>>> l = (('John', ['Hunter']), (1, 23), [[([42, (5, 23)], )]]) 

>>> print(list(flatten(l))) 

['John', 'Hunter', 1, 23, 42, 5, 23] 

 

By: Composite of Holger Krekel and Luther Blissett 

From: https://code.activestate.com/recipes/121294/ 

and Recipe 1.12 in cookbook 

""" 

for item in seq: 

if scalarp(item) or item is None: 

yield item 

else: 

for subitem in flatten(item, scalarp): 

yield subitem 

 

 

@deprecated('2.1', "sorted(..., key=itemgetter(...))") 

class Sorter(object): 

""" 

Sort by attribute or item 

 

Example usage:: 

 

sort = Sorter() 

 

list = [(1, 2), (4, 8), (0, 3)] 

dict = [{'a': 3, 'b': 4}, {'a': 5, 'b': 2}, {'a': 0, 'b': 0}, 

{'a': 9, 'b': 9}] 

 

 

sort(list) # default sort 

sort(list, 1) # sort by index 1 

sort(dict, 'a') # sort a list of dicts by key 'a' 

 

""" 

 

def _helper(self, data, aux, inplace): 

aux.sort() 

result = [data[i] for junk, i in aux] 

if inplace: 

data[:] = result 

return result 

 

def byItem(self, data, itemindex=None, inplace=1): 

if itemindex is None: 

if inplace: 

data.sort() 

result = data 

else: 

result = sorted(data) 

return result 

else: 

aux = [(data[i][itemindex], i) for i in range(len(data))] 

return self._helper(data, aux, inplace) 

 

def byAttribute(self, data, attributename, inplace=1): 

aux = [(getattr(data[i], attributename), i) for i in range(len(data))] 

return self._helper(data, aux, inplace) 

 

# a couple of handy synonyms 

sort = byItem 

__call__ = byItem 

 

 

@deprecated('2.1') 

class Xlator(dict): 

""" 

All-in-one multiple-string-substitution class 

 

Example usage:: 

 

text = "Larry Wall is the creator of Perl" 

adict = { 

"Larry Wall" : "Guido van Rossum", 

"creator" : "Benevolent Dictator for Life", 

"Perl" : "Python", 

} 

 

print(multiple_replace(adict, text)) 

 

xlat = Xlator(adict) 

print(xlat.xlat(text)) 

""" 

 

def _make_regex(self): 

""" Build re object based on the keys of the current dictionary """ 

return re.compile("|".join(map(re.escape, self))) 

 

def __call__(self, match): 

""" Handler invoked for each regex *match* """ 

return self[match.group(0)] 

 

def xlat(self, text): 

""" Translate *text*, returns the modified text. """ 

return self._make_regex().sub(self, text) 

 

 

@deprecated('2.1') 

def soundex(name, len=4): 

""" soundex module conforming to Odell-Russell algorithm """ 

 

# digits holds the soundex values for the alphabet 

soundex_digits = '01230120022455012623010202' 

sndx = '' 

fc = '' 

 

# Translate letters in name to soundex digits 

for c in name.upper(): 

if c.isalpha(): 

if not fc: 

fc = c # Remember first letter 

d = soundex_digits[ord(c) - ord('A')] 

# Duplicate consecutive soundex digits are skipped 

if not sndx or (d != sndx[-1]): 

sndx += d 

 

# Replace first digit with first letter 

sndx = fc + sndx[1:] 

 

# Remove all 0s from the soundex code 

sndx = sndx.replace('0', '') 

 

# Return soundex code truncated or 0-padded to len characters 

return (sndx + (len * '0'))[:len] 

 

 

@deprecated('2.1') 

class Null(object): 

""" Null objects always and reliably "do nothing." """ 

 

def __init__(self, *args, **kwargs): 

pass 

 

def __call__(self, *args, **kwargs): 

return self 

 

def __str__(self): 

return "Null()" 

 

def __repr__(self): 

return "Null()" 

 

if six.PY3: 

def __bool__(self): 

return 0 

else: 

def __nonzero__(self): 

return 0 

 

def __getattr__(self, name): 

return self 

 

def __setattr__(self, name, value): 

return self 

 

def __delattr__(self, name): 

return self 

 

 

def mkdirs(newdir, mode=0o777): 

""" 

make directory *newdir* recursively, and set *mode*. Equivalent to :: 

 

> mkdir -p NEWDIR 

> chmod MODE NEWDIR 

""" 

# this functionality is now in core python as of 3.2 

# LPY DROP 

if six.PY3: 

os.makedirs(newdir, mode=mode, exist_ok=True) 

else: 

try: 

os.makedirs(newdir, mode=mode) 

except OSError as exception: 

if exception.errno != errno.EEXIST: 

raise 

 

 

class GetRealpathAndStat(object): 

def __init__(self): 

self._cache = {} 

 

def __call__(self, path): 

result = self._cache.get(path) 

if result is None: 

realpath = os.path.realpath(path) 

if sys.platform == 'win32': 

stat_key = realpath 

else: 

stat = os.stat(realpath) 

stat_key = (stat.st_ino, stat.st_dev) 

result = realpath, stat_key 

self._cache[path] = result 

return result 

 

 

get_realpath_and_stat = GetRealpathAndStat() 

 

 

@deprecated('2.1') 

def dict_delall(d, keys): 

"""delete all of the *keys* from the :class:`dict` *d*""" 

for key in keys: 

try: 

del d[key] 

except KeyError: 

pass 

 

 

@deprecated('2.1') 

class RingBuffer(object): 

""" class that implements a not-yet-full buffer """ 

def __init__(self, size_max): 

self.max = size_max 

self.data = [] 

 

class __Full: 

""" class that implements a full buffer """ 

def append(self, x): 

""" Append an element overwriting the oldest one. """ 

self.data[self.cur] = x 

self.cur = (self.cur + 1) % self.max 

 

def get(self): 

""" return list of elements in correct order """ 

return self.data[self.cur:] + self.data[:self.cur] 

 

def append(self, x): 

"""append an element at the end of the buffer""" 

self.data.append(x) 

if len(self.data) == self.max: 

self.cur = 0 

# Permanently change self's class from non-full to full 

self.__class__ = __Full 

 

def get(self): 

""" Return a list of elements from the oldest to the newest. """ 

return self.data 

 

def __get_item__(self, i): 

return self.data[i % len(self.data)] 

 

 

@deprecated('2.1') 

def get_split_ind(seq, N): 

""" 

*seq* is a list of words. Return the index into seq such that:: 

 

len(' '.join(seq[:ind])<=N 

 

. 

""" 

 

s_len = 0 

# todo: use Alex's xrange pattern from the cbook for efficiency 

for (word, ind) in zip(seq, xrange(len(seq))): 

s_len += len(word) + 1 # +1 to account for the len(' ') 

if s_len >= N: 

return ind 

return len(seq) 

 

 

@deprecated('2.1', alternative='textwrap.TextWrapper') 

def wrap(prefix, text, cols): 

"""wrap *text* with *prefix* at length *cols*""" 

pad = ' ' * len(prefix.expandtabs()) 

available = cols - len(pad) 

 

seq = text.split(' ') 

Nseq = len(seq) 

ind = 0 

lines = [] 

while ind < Nseq: 

lastInd = ind 

ind += get_split_ind(seq[ind:], available) 

lines.append(seq[lastInd:ind]) 

 

# add the prefix to the first line, pad with spaces otherwise 

ret = prefix + ' '.join(lines[0]) + '\n' 

for line in lines[1:]: 

ret += pad + ' '.join(line) + '\n' 

return ret 

 

 

# A regular expression used to determine the amount of space to 

# remove. It looks for the first sequence of spaces immediately 

# following the first newline, or at the beginning of the string. 

_find_dedent_regex = re.compile(r"(?:(?:\n\r?)|^)( *)\S") 

# A cache to hold the regexs that actually remove the indent. 

_dedent_regex = {} 

 

 

def dedent(s): 

""" 

Remove excess indentation from docstring *s*. 

 

Discards any leading blank lines, then removes up to n whitespace 

characters from each line, where n is the number of leading 

whitespace characters in the first line. It differs from 

textwrap.dedent in its deletion of leading blank lines and its use 

of the first non-blank line to determine the indentation. 

 

It is also faster in most cases. 

""" 

# This implementation has a somewhat obtuse use of regular 

# expressions. However, this function accounted for almost 30% of 

# matplotlib startup time, so it is worthy of optimization at all 

# costs. 

 

if not s: # includes case of s is None 

return '' 

 

match = _find_dedent_regex.match(s) 

if match is None: 

return s 

 

# This is the number of spaces to remove from the left-hand side. 

nshift = match.end(1) - match.start(1) 

if nshift == 0: 

return s 

 

# Get a regex that will remove *up to* nshift spaces from the 

# beginning of each line. If it isn't in the cache, generate it. 

unindent = _dedent_regex.get(nshift, None) 

if unindent is None: 

unindent = re.compile("\n\r? {0,%d}" % nshift) 

_dedent_regex[nshift] = unindent 

 

result = unindent.sub("\n", s).strip() 

return result 

 

 

def listFiles(root, patterns='*', recurse=1, return_folders=0): 

""" 

Recursively list files 

 

from Parmar and Martelli in the Python Cookbook 

""" 

import os.path 

import fnmatch 

# Expand patterns from semicolon-separated string to list 

pattern_list = patterns.split(';') 

results = [] 

 

for dirname, dirs, files in os.walk(root): 

# Append to results all relevant files (and perhaps folders) 

for name in files: 

fullname = os.path.normpath(os.path.join(dirname, name)) 

if return_folders or os.path.isfile(fullname): 

for pattern in pattern_list: 

if fnmatch.fnmatch(name, pattern): 

results.append(fullname) 

break 

# Block recursion if recursion was disallowed 

if not recurse: 

break 

 

return results 

 

 

@deprecated('2.1') 

def get_recursive_filelist(args): 

""" 

Recurse all the files and dirs in *args* ignoring symbolic links 

and return the files as a list of strings 

""" 

files = [] 

 

for arg in args: 

if os.path.isfile(arg): 

files.append(arg) 

continue 

if os.path.isdir(arg): 

newfiles = listFiles(arg, recurse=1, return_folders=1) 

files.extend(newfiles) 

 

return [f for f in files if not os.path.islink(f)] 

 

 

@deprecated('2.1') 

def pieces(seq, num=2): 

"""Break up the *seq* into *num* tuples""" 

start = 0 

while 1: 

item = seq[start:start + num] 

if not len(item): 

break 

yield item 

start += num 

 

 

@deprecated('2.1') 

def exception_to_str(s=None): 

if six.PY3: 

sh = io.StringIO() 

else: 

sh = io.BytesIO() 

if s is not None: 

print(s, file=sh) 

traceback.print_exc(file=sh) 

return sh.getvalue() 

 

 

@deprecated('2.1') 

def allequal(seq): 

""" 

Return *True* if all elements of *seq* compare equal. If *seq* is 

0 or 1 length, return *True* 

""" 

if len(seq) < 2: 

return True 

val = seq[0] 

for i in xrange(1, len(seq)): 

thisval = seq[i] 

if thisval != val: 

return False 

return True 

 

 

@deprecated('2.1') 

def alltrue(seq): 

""" 

Return *True* if all elements of *seq* evaluate to *True*. If 

*seq* is empty, return *False*. 

""" 

if not len(seq): 

return False 

for val in seq: 

if not val: 

return False 

return True 

 

 

@deprecated('2.1') 

def onetrue(seq): 

""" 

Return *True* if one element of *seq* is *True*. It *seq* is 

empty, return *False*. 

""" 

if not len(seq): 

return False 

for val in seq: 

if val: 

return True 

return False 

 

 

@deprecated('2.1') 

def allpairs(x): 

""" 

return all possible pairs in sequence *x* 

""" 

return [(s, f) for i, f in enumerate(x) for s in x[i + 1:]] 

 

 

class maxdict(dict): 

""" 

A dictionary with a maximum size; this doesn't override all the 

relevant methods to constrain the size, just setitem, so use with 

caution 

""" 

def __init__(self, maxsize): 

dict.__init__(self) 

self.maxsize = maxsize 

self._killkeys = [] 

 

def __setitem__(self, k, v): 

if k not in self: 

if len(self) >= self.maxsize: 

del self[self._killkeys[0]] 

del self._killkeys[0] 

self._killkeys.append(k) 

dict.__setitem__(self, k, v) 

 

 

class Stack(object): 

""" 

Implement a stack where elements can be pushed on and you can move 

back and forth. But no pop. Should mimic home / back / forward 

in a browser 

""" 

 

def __init__(self, default=None): 

self.clear() 

self._default = default 

 

def __call__(self): 

"""return the current element, or None""" 

if not len(self._elements): 

return self._default 

else: 

return self._elements[self._pos] 

 

def __len__(self): 

return self._elements.__len__() 

 

def __getitem__(self, ind): 

return self._elements.__getitem__(ind) 

 

def forward(self): 

"""move the position forward and return the current element""" 

n = len(self._elements) 

if self._pos < n - 1: 

self._pos += 1 

return self() 

 

def back(self): 

"""move the position back and return the current element""" 

if self._pos > 0: 

self._pos -= 1 

return self() 

 

def push(self, o): 

""" 

push object onto stack at current position - all elements 

occurring later than the current position are discarded 

""" 

self._elements = self._elements[:self._pos + 1] 

self._elements.append(o) 

self._pos = len(self._elements) - 1 

return self() 

 

def home(self): 

"""push the first element onto the top of the stack""" 

if not len(self._elements): 

return 

self.push(self._elements[0]) 

return self() 

 

def empty(self): 

return len(self._elements) == 0 

 

def clear(self): 

"""empty the stack""" 

self._pos = -1 

self._elements = [] 

 

def bubble(self, o): 

""" 

raise *o* to the top of the stack and return *o*. *o* must be 

in the stack 

""" 

 

if o not in self._elements: 

raise ValueError('Unknown element o') 

old = self._elements[:] 

self.clear() 

bubbles = [] 

for thiso in old: 

if thiso == o: 

bubbles.append(thiso) 

else: 

self.push(thiso) 

for thiso in bubbles: 

self.push(o) 

return o 

 

def remove(self, o): 

'remove element *o* from the stack' 

if o not in self._elements: 

raise ValueError('Unknown element o') 

old = self._elements[:] 

self.clear() 

for thiso in old: 

if thiso == o: 

continue 

else: 

self.push(thiso) 

 

 

@deprecated('2.1') 

def finddir(o, match, case=False): 

""" 

return all attributes of *o* which match string in match. if case 

is True require an exact case match. 

""" 

if case: 

names = [(name, name) for name in dir(o) 

if isinstance(name, six.string_types)] 

else: 

names = [(name.lower(), name) for name in dir(o) 

if isinstance(name, six.string_types)] 

match = match.lower() 

return [orig for name, orig in names if name.find(match) >= 0] 

 

 

@deprecated('2.1') 

def reverse_dict(d): 

"""reverse the dictionary -- may lose data if values are not unique!""" 

return {v: k for k, v in six.iteritems(d)} 

 

 

@deprecated('2.1') 

def restrict_dict(d, keys): 

""" 

Return a dictionary that contains those keys that appear in both 

d and keys, with values from d. 

""" 

return {k: v for k, v in six.iteritems(d) if k in keys} 

 

 

def report_memory(i=0): # argument may go away 

"""return the memory consumed by process""" 

from matplotlib.compat.subprocess import Popen, PIPE 

pid = os.getpid() 

if sys.platform == 'sunos5': 

try: 

a2 = Popen(str('ps -p %d -o osz') % pid, shell=True, 

stdout=PIPE).stdout.readlines() 

except OSError: 

raise NotImplementedError( 

"report_memory works on Sun OS only if " 

"the 'ps' program is found") 

mem = int(a2[-1].strip()) 

elif sys.platform.startswith('linux'): 

try: 

a2 = Popen(str('ps -p %d -o rss,sz') % pid, shell=True, 

stdout=PIPE).stdout.readlines() 

except OSError: 

raise NotImplementedError( 

"report_memory works on Linux only if " 

"the 'ps' program is found") 

mem = int(a2[1].split()[1]) 

elif sys.platform.startswith('darwin'): 

try: 

a2 = Popen(str('ps -p %d -o rss,vsz') % pid, shell=True, 

stdout=PIPE).stdout.readlines() 

except OSError: 

raise NotImplementedError( 

"report_memory works on Mac OS only if " 

"the 'ps' program is found") 

mem = int(a2[1].split()[0]) 

elif sys.platform.startswith('win'): 

try: 

a2 = Popen([str("tasklist"), "/nh", "/fi", "pid eq %d" % pid], 

stdout=PIPE).stdout.read() 

except OSError: 

raise NotImplementedError( 

"report_memory works on Windows only if " 

"the 'tasklist' program is found") 

mem = int(a2.strip().split()[-2].replace(',', '')) 

else: 

raise NotImplementedError( 

"We don't have a memory monitor for %s" % sys.platform) 

return mem 

 

 

_safezip_msg = 'In safezip, len(args[0])=%d but len(args[%d])=%d' 

 

 

def safezip(*args): 

"""make sure *args* are equal len before zipping""" 

Nx = len(args[0]) 

for i, arg in enumerate(args[1:]): 

if len(arg) != Nx: 

raise ValueError(_safezip_msg % (Nx, i + 1, len(arg))) 

return list(zip(*args)) 

 

 

@deprecated('2.1') 

def issubclass_safe(x, klass): 

"""return issubclass(x, klass) and return False on a TypeError""" 

 

try: 

return issubclass(x, klass) 

except TypeError: 

return False 

 

 

def safe_masked_invalid(x, copy=False): 

x = np.array(x, subok=True, copy=copy) 

if not x.dtype.isnative: 

# Note that the argument to `byteswap` is 'inplace', 

# thus if we have already made a copy, do the byteswap in 

# place, else make a copy with the byte order swapped. 

# Be explicit that we are swapping the byte order of the dtype 

x = x.byteswap(copy).newbyteorder('S') 

 

try: 

xm = np.ma.masked_invalid(x, copy=False) 

xm.shrink_mask() 

except TypeError: 

return x 

return xm 

 

 

def print_cycles(objects, outstream=sys.stdout, show_progress=False): 

""" 

*objects* 

A list of objects to find cycles in. It is often useful to 

pass in gc.garbage to find the cycles that are preventing some 

objects from being garbage collected. 

 

*outstream* 

The stream for output. 

 

*show_progress* 

If True, print the number of objects reached as they are found. 

""" 

import gc 

from types import FrameType 

 

def print_path(path): 

for i, step in enumerate(path): 

# next "wraps around" 

next = path[(i + 1) % len(path)] 

 

outstream.write(" %s -- " % str(type(step))) 

if isinstance(step, dict): 

for key, val in six.iteritems(step): 

if val is next: 

outstream.write("[%s]" % repr(key)) 

break 

if key is next: 

outstream.write("[key] = %s" % repr(val)) 

break 

elif isinstance(step, list): 

outstream.write("[%d]" % step.index(next)) 

elif isinstance(step, tuple): 

outstream.write("( tuple )") 

else: 

outstream.write(repr(step)) 

outstream.write(" ->\n") 

outstream.write("\n") 

 

def recurse(obj, start, all, current_path): 

if show_progress: 

outstream.write("%d\r" % len(all)) 

 

all[id(obj)] = None 

 

referents = gc.get_referents(obj) 

for referent in referents: 

# If we've found our way back to the start, this is 

# a cycle, so print it out 

if referent is start: 

print_path(current_path) 

 

# Don't go back through the original list of objects, or 

# through temporary references to the object, since those 

# are just an artifact of the cycle detector itself. 

elif referent is objects or isinstance(referent, FrameType): 

continue 

 

# We haven't seen this object before, so recurse 

elif id(referent) not in all: 

recurse(referent, start, all, current_path + [obj]) 

 

for obj in objects: 

outstream.write("Examining: %r\n" % (obj,)) 

recurse(obj, obj, {}, []) 

 

 

class Grouper(object): 

""" 

This class provides a lightweight way to group arbitrary objects 

together into disjoint sets when a full-blown graph data structure 

would be overkill. 

 

Objects can be joined using :meth:`join`, tested for connectedness 

using :meth:`joined`, and all disjoint sets can be retrieved by 

using the object as an iterator. 

 

The objects being joined must be hashable and weak-referenceable. 

 

For example: 

 

>>> from matplotlib.cbook import Grouper 

>>> class Foo(object): 

... def __init__(self, s): 

... self.s = s 

... def __repr__(self): 

... return self.s 

... 

>>> a, b, c, d, e, f = [Foo(x) for x in 'abcdef'] 

>>> grp = Grouper() 

>>> grp.join(a, b) 

>>> grp.join(b, c) 

>>> grp.join(d, e) 

>>> sorted(map(tuple, grp)) 

[(a, b, c), (d, e)] 

>>> grp.joined(a, b) 

True 

>>> grp.joined(a, c) 

True 

>>> grp.joined(a, d) 

False 

 

""" 

def __init__(self, init=()): 

mapping = self._mapping = {} 

for x in init: 

mapping[ref(x)] = [ref(x)] 

 

def __contains__(self, item): 

return ref(item) in self._mapping 

 

def clean(self): 

""" 

Clean dead weak references from the dictionary 

""" 

mapping = self._mapping 

to_drop = [key for key in mapping if key() is None] 

for key in to_drop: 

val = mapping.pop(key) 

val.remove(key) 

 

def join(self, a, *args): 

""" 

Join given arguments into the same set. Accepts one or more 

arguments. 

""" 

mapping = self._mapping 

set_a = mapping.setdefault(ref(a), [ref(a)]) 

 

for arg in args: 

set_b = mapping.get(ref(arg)) 

if set_b is None: 

set_a.append(ref(arg)) 

mapping[ref(arg)] = set_a 

elif set_b is not set_a: 

if len(set_b) > len(set_a): 

set_a, set_b = set_b, set_a 

set_a.extend(set_b) 

for elem in set_b: 

mapping[elem] = set_a 

 

self.clean() 

 

def joined(self, a, b): 

""" 

Returns True if *a* and *b* are members of the same set. 

""" 

self.clean() 

 

mapping = self._mapping 

try: 

return mapping[ref(a)] is mapping[ref(b)] 

except KeyError: 

return False 

 

def remove(self, a): 

self.clean() 

 

mapping = self._mapping 

seta = mapping.pop(ref(a), None) 

if seta is not None: 

seta.remove(ref(a)) 

 

def __iter__(self): 

""" 

Iterate over each of the disjoint sets as a list. 

 

The iterator is invalid if interleaved with calls to join(). 

""" 

self.clean() 

token = object() 

 

# Mark each group as we come across if by appending a token, 

# and don't yield it twice 

for group in six.itervalues(self._mapping): 

if group[-1] is not token: 

yield [x() for x in group] 

group.append(token) 

 

# Cleanup the tokens 

for group in six.itervalues(self._mapping): 

if group[-1] is token: 

del group[-1] 

 

def get_siblings(self, a): 

""" 

Returns all of the items joined with *a*, including itself. 

""" 

self.clean() 

 

siblings = self._mapping.get(ref(a), [ref(a)]) 

return [x() for x in siblings] 

 

 

def simple_linear_interpolation(a, steps): 

""" 

Resample an array with ``steps - 1`` points between original point pairs. 

 

Parameters 

---------- 

a : array, shape (n, ...) 

steps : int 

 

Returns 

------- 

array, shape ``((n - 1) * steps + 1, ...)`` 

 

Along each column of *a*, ``(steps - 1)`` points are introduced between 

each original values; the values are linearly interpolated. 

""" 

fps = a.reshape((len(a), -1)) 

xp = np.arange(len(a)) * steps 

x = np.arange((len(a) - 1) * steps + 1) 

return (np.column_stack([np.interp(x, xp, fp) for fp in fps.T]) 

.reshape((len(x),) + a.shape[1:])) 

 

 

@deprecated('2.1', alternative='shutil.rmtree') 

def recursive_remove(path): 

if os.path.isdir(path): 

for fname in (glob.glob(os.path.join(path, '*')) + 

glob.glob(os.path.join(path, '.*'))): 

if os.path.isdir(fname): 

recursive_remove(fname) 

os.removedirs(fname) 

else: 

os.remove(fname) 

# os.removedirs(path) 

else: 

os.remove(path) 

 

 

def delete_masked_points(*args): 

""" 

Find all masked and/or non-finite points in a set of arguments, 

and return the arguments with only the unmasked points remaining. 

 

Arguments can be in any of 5 categories: 

 

1) 1-D masked arrays 

2) 1-D ndarrays 

3) ndarrays with more than one dimension 

4) other non-string iterables 

5) anything else 

 

The first argument must be in one of the first four categories; 

any argument with a length differing from that of the first 

argument (and hence anything in category 5) then will be 

passed through unchanged. 

 

Masks are obtained from all arguments of the correct length 

in categories 1, 2, and 4; a point is bad if masked in a masked 

array or if it is a nan or inf. No attempt is made to 

extract a mask from categories 2, 3, and 4 if :meth:`np.isfinite` 

does not yield a Boolean array. 

 

All input arguments that are not passed unchanged are returned 

as ndarrays after removing the points or rows corresponding to 

masks in any of the arguments. 

 

A vastly simpler version of this function was originally 

written as a helper for Axes.scatter(). 

 

""" 

if not len(args): 

return () 

if (isinstance(args[0], six.string_types) or not iterable(args[0])): 

raise ValueError("First argument must be a sequence") 

nrecs = len(args[0]) 

margs = [] 

seqlist = [False] * len(args) 

for i, x in enumerate(args): 

if (not isinstance(x, six.string_types) and iterable(x) 

and len(x) == nrecs): 

seqlist[i] = True 

if isinstance(x, np.ma.MaskedArray): 

if x.ndim > 1: 

raise ValueError("Masked arrays must be 1-D") 

else: 

x = np.asarray(x) 

margs.append(x) 

masks = [] # list of masks that are True where good 

for i, x in enumerate(margs): 

if seqlist[i]: 

if x.ndim > 1: 

continue # Don't try to get nan locations unless 1-D. 

if isinstance(x, np.ma.MaskedArray): 

masks.append(~np.ma.getmaskarray(x)) # invert the mask 

xd = x.data 

else: 

xd = x 

try: 

mask = np.isfinite(xd) 

if isinstance(mask, np.ndarray): 

masks.append(mask) 

except: # Fixme: put in tuple of possible exceptions? 

pass 

if len(masks): 

mask = np.logical_and.reduce(masks) 

igood = mask.nonzero()[0] 

if len(igood) < nrecs: 

for i, x in enumerate(margs): 

if seqlist[i]: 

margs[i] = x.take(igood, axis=0) 

for i, x in enumerate(margs): 

if seqlist[i] and isinstance(x, np.ma.MaskedArray): 

margs[i] = x.filled() 

return margs 

 

 

def boxplot_stats(X, whis=1.5, bootstrap=None, labels=None, 

autorange=False): 

""" 

Returns list of dictionaries of statistics used to draw a series 

of box and whisker plots. The `Returns` section enumerates the 

required keys of the dictionary. Users can skip this function and 

pass a user-defined set of dictionaries to the new `axes.bxp` method 

instead of relying on MPL to do the calculations. 

 

Parameters 

---------- 

X : array-like 

Data that will be represented in the boxplots. Should have 2 or 

fewer dimensions. 

 

whis : float, string, or sequence (default = 1.5) 

As a float, determines the reach of the whiskers to the beyond the 

first and third quartiles. In other words, where IQR is the 

interquartile range (`Q3-Q1`), the upper whisker will extend to last 

datum less than `Q3 + whis*IQR`). Similarly, the lower whisker will 

extend to the first datum greater than `Q1 - whis*IQR`. 

Beyond the whiskers, data are considered outliers 

and are plotted as individual points. This can be set this to an 

ascending sequence of percentile (e.g., [5, 95]) to set the 

whiskers at specific percentiles of the data. Finally, `whis` 

can be the string ``'range'`` to force the whiskers to the 

minimum and maximum of the data. In the edge case that the 25th 

and 75th percentiles are equivalent, `whis` can be automatically 

set to ``'range'`` via the `autorange` option. 

 

bootstrap : int, optional 

Number of times the confidence intervals around the median 

should be bootstrapped (percentile method). 

 

labels : array-like, optional 

Labels for each dataset. Length must be compatible with 

dimensions of `X`. 

 

autorange : bool, optional (False) 

When `True` and the data are distributed such that the 25th and 

75th percentiles are equal, ``whis`` is set to ``'range'`` such 

that the whisker ends are at the minimum and maximum of the 

data. 

 

Returns 

------- 

bxpstats : list of dict 

A list of dictionaries containing the results for each column 

of data. Keys of each dictionary are the following: 

 

======== =================================== 

Key Value Description 

======== =================================== 

label tick label for the boxplot 

mean arithemetic mean value 

med 50th percentile 

q1 first quartile (25th percentile) 

q3 third quartile (75th percentile) 

cilo lower notch around the median 

cihi upper notch around the median 

whislo end of the lower whisker 

whishi end of the upper whisker 

fliers outliers 

======== =================================== 

 

Notes 

----- 

Non-bootstrapping approach to confidence interval uses Gaussian- 

based asymptotic approximation: 

 

.. math:: 

 

\\mathrm{med} \\pm 1.57 \\times \\frac{\\mathrm{iqr}}{\\sqrt{N}} 

 

General approach from: 

McGill, R., Tukey, J.W., and Larsen, W.A. (1978) "Variations of 

Boxplots", The American Statistician, 32:12-16. 

 

""" 

 

def _bootstrap_median(data, N=5000): 

# determine 95% confidence intervals of the median 

M = len(data) 

percentiles = [2.5, 97.5] 

 

bs_index = np.random.randint(M, size=(N, M)) 

bsData = data[bs_index] 

estimate = np.median(bsData, axis=1, overwrite_input=True) 

 

CI = np.percentile(estimate, percentiles) 

return CI 

 

def _compute_conf_interval(data, med, iqr, bootstrap): 

if bootstrap is not None: 

# Do a bootstrap estimate of notch locations. 

# get conf. intervals around median 

CI = _bootstrap_median(data, N=bootstrap) 

notch_min = CI[0] 

notch_max = CI[1] 

else: 

 

N = len(data) 

notch_min = med - 1.57 * iqr / np.sqrt(N) 

notch_max = med + 1.57 * iqr / np.sqrt(N) 

 

return notch_min, notch_max 

 

# output is a list of dicts 

bxpstats = [] 

 

# convert X to a list of lists 

X = _reshape_2D(X, "X") 

 

ncols = len(X) 

if labels is None: 

labels = repeat(None) 

elif len(labels) != ncols: 

raise ValueError("Dimensions of labels and X must be compatible") 

 

input_whis = whis 

for ii, (x, label) in enumerate(zip(X, labels), start=0): 

 

# empty dict 

stats = {} 

if label is not None: 

stats['label'] = label 

 

# restore whis to the input values in case it got changed in the loop 

whis = input_whis 

 

# note tricksyness, append up here and then mutate below 

bxpstats.append(stats) 

 

# if empty, bail 

if len(x) == 0: 

stats['fliers'] = np.array([]) 

stats['mean'] = np.nan 

stats['med'] = np.nan 

stats['q1'] = np.nan 

stats['q3'] = np.nan 

stats['cilo'] = np.nan 

stats['cihi'] = np.nan 

stats['whislo'] = np.nan 

stats['whishi'] = np.nan 

stats['med'] = np.nan 

continue 

 

# up-convert to an array, just to be safe 

x = np.asarray(x) 

 

# arithmetic mean 

stats['mean'] = np.mean(x) 

 

# medians and quartiles 

q1, med, q3 = np.percentile(x, [25, 50, 75]) 

 

# interquartile range 

stats['iqr'] = q3 - q1 

if stats['iqr'] == 0 and autorange: 

whis = 'range' 

 

# conf. interval around median 

stats['cilo'], stats['cihi'] = _compute_conf_interval( 

x, med, stats['iqr'], bootstrap 

) 

 

# lowest/highest non-outliers 

if np.isscalar(whis): 

if np.isreal(whis): 

loval = q1 - whis * stats['iqr'] 

hival = q3 + whis * stats['iqr'] 

elif whis in ['range', 'limit', 'limits', 'min/max']: 

loval = np.min(x) 

hival = np.max(x) 

else: 

raise ValueError('whis must be a float, valid string, or list ' 

'of percentiles') 

else: 

loval = np.percentile(x, whis[0]) 

hival = np.percentile(x, whis[1]) 

 

# get high extreme 

wiskhi = np.compress(x <= hival, x) 

if len(wiskhi) == 0 or np.max(wiskhi) < q3: 

stats['whishi'] = q3 

else: 

stats['whishi'] = np.max(wiskhi) 

 

# get low extreme 

wisklo = np.compress(x >= loval, x) 

if len(wisklo) == 0 or np.min(wisklo) > q1: 

stats['whislo'] = q1 

else: 

stats['whislo'] = np.min(wisklo) 

 

# compute a single array of outliers 

stats['fliers'] = np.hstack([ 

np.compress(x < stats['whislo'], x), 

np.compress(x > stats['whishi'], x) 

]) 

 

# add in the remaining stats 

stats['q1'], stats['med'], stats['q3'] = q1, med, q3 

 

return bxpstats 

 

 

# FIXME I don't think this is used anywhere 

@deprecated('2.1') 

def unmasked_index_ranges(mask, compressed=True): 

""" 

Find index ranges where *mask* is *False*. 

 

*mask* will be flattened if it is not already 1-D. 

 

Returns Nx2 :class:`numpy.ndarray` with each row the start and stop 

indices for slices of the compressed :class:`numpy.ndarray` 

corresponding to each of *N* uninterrupted runs of unmasked 

values. If optional argument *compressed* is *False*, it returns 

the start and stop indices into the original :class:`numpy.ndarray`, 

not the compressed :class:`numpy.ndarray`. Returns *None* if there 

are no unmasked values. 

 

Example:: 

 

y = ma.array(np.arange(5), mask = [0,0,1,0,0]) 

ii = unmasked_index_ranges(ma.getmaskarray(y)) 

# returns array [[0,2,] [2,4,]] 

 

y.compressed()[ii[1,0]:ii[1,1]] 

# returns array [3,4,] 

 

ii = unmasked_index_ranges(ma.getmaskarray(y), compressed=False) 

# returns array [[0, 2], [3, 5]] 

 

y.filled()[ii[1,0]:ii[1,1]] 

# returns array [3,4,] 

 

Prior to the transforms refactoring, this was used to support 

masked arrays in Line2D. 

""" 

mask = mask.reshape(mask.size) 

m = np.concatenate(((1,), mask, (1,))) 

indices = np.arange(len(mask) + 1) 

mdif = m[1:] - m[:-1] 

i0 = np.compress(mdif == -1, indices) 

i1 = np.compress(mdif == 1, indices) 

assert len(i0) == len(i1) 

if len(i1) == 0: 

return None # Maybe this should be np.zeros((0,2), dtype=int) 

if not compressed: 

return np.concatenate((i0[:, np.newaxis], i1[:, np.newaxis]), axis=1) 

seglengths = i1 - i0 

breakpoints = np.cumsum(seglengths) 

ic0 = np.concatenate(((0,), breakpoints[:-1])) 

ic1 = breakpoints 

return np.concatenate((ic0[:, np.newaxis], ic1[:, np.newaxis]), axis=1) 

 

 

# The ls_mapper maps short codes for line style to their full name used by 

# backends; the reverse mapper is for mapping full names to short ones. 

ls_mapper = {'-': 'solid', '--': 'dashed', '-.': 'dashdot', ':': 'dotted'} 

ls_mapper_r = {v: k for k, v in six.iteritems(ls_mapper)} 

 

 

@deprecated('2.2') 

def align_iterators(func, *iterables): 

""" 

This generator takes a bunch of iterables that are ordered by func 

It sends out ordered tuples:: 

 

(func(row), [rows from all iterators matching func(row)]) 

 

It is used by :func:`matplotlib.mlab.recs_join` to join record arrays 

""" 

class myiter: 

def __init__(self, it): 

self.it = it 

self.key = self.value = None 

self.iternext() 

 

def iternext(self): 

try: 

self.value = next(self.it) 

self.key = func(self.value) 

except StopIteration: 

self.value = self.key = None 

 

def __call__(self, key): 

retval = None 

if key == self.key: 

retval = self.value 

self.iternext() 

elif self.key and key > self.key: 

raise ValueError("Iterator has been left behind") 

return retval 

 

# This can be made more efficient by not computing the minimum key for each 

# iteration 

iters = [myiter(it) for it in iterables] 

minvals = minkey = True 

while True: 

minvals = ([_f for _f in [it.key for it in iters] if _f]) 

if minvals: 

minkey = min(minvals) 

yield (minkey, [it(minkey) for it in iters]) 

else: 

break 

 

 

def contiguous_regions(mask): 

""" 

Return a list of (ind0, ind1) such that mask[ind0:ind1].all() is 

True and we cover all such regions 

""" 

mask = np.asarray(mask, dtype=bool) 

 

if not mask.size: 

return [] 

 

# Find the indices of region changes, and correct offset 

idx, = np.nonzero(mask[:-1] != mask[1:]) 

idx += 1 

 

# List operations are faster for moderately sized arrays 

idx = idx.tolist() 

 

# Add first and/or last index if needed 

if mask[0]: 

idx = [0] + idx 

if mask[-1]: 

idx.append(len(mask)) 

 

return list(zip(idx[::2], idx[1::2])) 

 

 

def is_math_text(s): 

# Did we find an even number of non-escaped dollar signs? 

# If so, treat is as math text. 

try: 

s = six.text_type(s) 

except UnicodeDecodeError: 

raise ValueError( 

"matplotlib display text must have all code points < 128 or use " 

"Unicode strings") 

 

dollar_count = s.count(r'$') - s.count(r'\$') 

even_dollars = (dollar_count > 0 and dollar_count % 2 == 0) 

 

return even_dollars 

 

 

def _to_unmasked_float_array(x): 

""" 

Convert a sequence to a float array; if input was a masked array, masked 

values are converted to nans. 

""" 

if hasattr(x, 'mask'): 

return np.ma.asarray(x, float).filled(np.nan) 

else: 

return np.asarray(x, float) 

 

 

def _check_1d(x): 

''' 

Converts a sequence of less than 1 dimension, to an array of 1 

dimension; leaves everything else untouched. 

''' 

if not hasattr(x, 'shape') or len(x.shape) < 1: 

return np.atleast_1d(x) 

else: 

try: 

x[:, None] 

return x 

except (IndexError, TypeError): 

return np.atleast_1d(x) 

 

 

def _reshape_2D(X, name): 

""" 

Use Fortran ordering to convert ndarrays and lists of iterables to lists of 

1D arrays. 

 

Lists of iterables are converted by applying `np.asarray` to each of their 

elements. 1D ndarrays are returned in a singleton list containing them. 

2D ndarrays are converted to the list of their *columns*. 

 

*name* is used to generate the error message for invalid inputs. 

""" 

# Iterate over columns for ndarrays, over rows otherwise. 

X = np.atleast_1d(X.T if isinstance(X, np.ndarray) else np.asarray(X)) 

if X.ndim == 1 and X.dtype.type != np.object_: 

# 1D array of scalars: directly return it. 

return [X] 

elif X.ndim in [1, 2]: 

# 2D array, or 1D array of iterables: flatten them first. 

return [np.reshape(x, -1) for x in X] 

else: 

raise ValueError("{} must have 2 or fewer dimensions".format(name)) 

 

 

def violin_stats(X, method, points=100): 

""" 

Returns a list of dictionaries of data which can be used to draw a series 

of violin plots. See the `Returns` section below to view the required keys 

of the dictionary. Users can skip this function and pass a user-defined set 

of dictionaries to the `axes.vplot` method instead of using MPL to do the 

calculations. 

 

Parameters 

---------- 

X : array-like 

Sample data that will be used to produce the gaussian kernel density 

estimates. Must have 2 or fewer dimensions. 

 

method : callable 

The method used to calculate the kernel density estimate for each 

column of data. When called via `method(v, coords)`, it should 

return a vector of the values of the KDE evaluated at the values 

specified in coords. 

 

points : scalar, default = 100 

Defines the number of points to evaluate each of the gaussian kernel 

density estimates at. 

 

Returns 

------- 

 

A list of dictionaries containing the results for each column of data. 

The dictionaries contain at least the following: 

 

- coords: A list of scalars containing the coordinates this particular 

kernel density estimate was evaluated at. 

- vals: A list of scalars containing the values of the kernel density 

estimate at each of the coordinates given in `coords`. 

- mean: The mean value for this column of data. 

- median: The median value for this column of data. 

- min: The minimum value for this column of data. 

- max: The maximum value for this column of data. 

""" 

 

# List of dictionaries describing each of the violins. 

vpstats = [] 

 

# Want X to be a list of data sequences 

X = _reshape_2D(X, "X") 

 

for x in X: 

# Dictionary of results for this distribution 

stats = {} 

 

# Calculate basic stats for the distribution 

min_val = np.min(x) 

max_val = np.max(x) 

 

# Evaluate the kernel density estimate 

coords = np.linspace(min_val, max_val, points) 

stats['vals'] = method(x, coords) 

stats['coords'] = coords 

 

# Store additional statistics for this distribution 

stats['mean'] = np.mean(x) 

stats['median'] = np.median(x) 

stats['min'] = min_val 

stats['max'] = max_val 

 

# Append to output 

vpstats.append(stats) 

 

return vpstats 

 

 

class _NestedClassGetter(object): 

# recipe from http://stackoverflow.com/a/11493777/741316 

""" 

When called with the containing class as the first argument, 

and the name of the nested class as the second argument, 

returns an instance of the nested class. 

""" 

def __call__(self, containing_class, class_name): 

nested_class = getattr(containing_class, class_name) 

 

# make an instance of a simple object (this one will do), for which we 

# can change the __class__ later on. 

nested_instance = _NestedClassGetter() 

 

# set the class of the instance, the __init__ will never be called on 

# the class but the original state will be set later on by pickle. 

nested_instance.__class__ = nested_class 

return nested_instance 

 

 

class _InstanceMethodPickler(object): 

""" 

Pickle cannot handle instancemethod saving. _InstanceMethodPickler 

provides a solution to this. 

""" 

def __init__(self, instancemethod): 

"""Takes an instancemethod as its only argument.""" 

if six.PY3: 

self.parent_obj = instancemethod.__self__ 

self.instancemethod_name = instancemethod.__func__.__name__ 

else: 

self.parent_obj = instancemethod.im_self 

self.instancemethod_name = instancemethod.im_func.__name__ 

 

def get_instancemethod(self): 

return getattr(self.parent_obj, self.instancemethod_name) 

 

 

def pts_to_prestep(x, *args): 

""" 

Convert continuous line to pre-steps. 

 

Given a set of ``N`` points, convert to ``2N - 1`` points, which when 

connected linearly give a step function which changes values at the 

beginning of the intervals. 

 

Parameters 

---------- 

x : array 

The x location of the steps. May be empty. 

 

y1, ..., yp : array 

y arrays to be turned into steps; all must be the same length as ``x``. 

 

Returns 

------- 

out : array 

The x and y values converted to steps in the same order as the input; 

can be unpacked as ``x_out, y1_out, ..., yp_out``. If the input is 

length ``N``, each of these arrays will be length ``2N + 1``. For 

``N=0``, the length will be 0. 

 

Examples 

-------- 

>> x_s, y1_s, y2_s = pts_to_prestep(x, y1, y2) 

""" 

steps = np.zeros((1 + len(args), max(2 * len(x) - 1, 0))) 

# In all `pts_to_*step` functions, only assign *once* using `x` and `args`, 

# as converting to an array may be expensive. 

steps[0, 0::2] = x 

steps[0, 1::2] = steps[0, 0:-2:2] 

steps[1:, 0::2] = args 

steps[1:, 1::2] = steps[1:, 2::2] 

return steps 

 

 

def pts_to_poststep(x, *args): 

""" 

Convert continuous line to post-steps. 

 

Given a set of ``N`` points convert to ``2N + 1`` points, which when 

connected linearly give a step function which changes values at the end of 

the intervals. 

 

Parameters 

---------- 

x : array 

The x location of the steps. May be empty. 

 

y1, ..., yp : array 

y arrays to be turned into steps; all must be the same length as ``x``. 

 

Returns 

------- 

out : array 

The x and y values converted to steps in the same order as the input; 

can be unpacked as ``x_out, y1_out, ..., yp_out``. If the input is 

length ``N``, each of these arrays will be length ``2N + 1``. For 

``N=0``, the length will be 0. 

 

Examples 

-------- 

>> x_s, y1_s, y2_s = pts_to_poststep(x, y1, y2) 

""" 

steps = np.zeros((1 + len(args), max(2 * len(x) - 1, 0))) 

steps[0, 0::2] = x 

steps[0, 1::2] = steps[0, 2::2] 

steps[1:, 0::2] = args 

steps[1:, 1::2] = steps[1:, 0:-2:2] 

return steps 

 

 

def pts_to_midstep(x, *args): 

""" 

Convert continuous line to mid-steps. 

 

Given a set of ``N`` points convert to ``2N`` points which when connected 

linearly give a step function which changes values at the middle of the 

intervals. 

 

Parameters 

---------- 

x : array 

The x location of the steps. May be empty. 

 

y1, ..., yp : array 

y arrays to be turned into steps; all must be the same length as ``x``. 

 

Returns 

------- 

out : array 

The x and y values converted to steps in the same order as the input; 

can be unpacked as ``x_out, y1_out, ..., yp_out``. If the input is 

length ``N``, each of these arrays will be length ``2N``. 

 

Examples 

-------- 

>> x_s, y1_s, y2_s = pts_to_midstep(x, y1, y2) 

""" 

steps = np.zeros((1 + len(args), 2 * len(x))) 

x = np.asanyarray(x) 

steps[0, 1:-1:2] = steps[0, 2::2] = (x[:-1] + x[1:]) / 2 

steps[0, :1] = x[:1] # Also works for zero-sized input. 

steps[0, -1:] = x[-1:] 

steps[1:, 0::2] = args 

steps[1:, 1::2] = steps[1:, 0::2] 

return steps 

 

 

STEP_LOOKUP_MAP = {'default': lambda x, y: (x, y), 

'steps': pts_to_prestep, 

'steps-pre': pts_to_prestep, 

'steps-post': pts_to_poststep, 

'steps-mid': pts_to_midstep} 

 

 

def index_of(y): 

""" 

A helper function to get the index of an input to plot 

against if x values are not explicitly given. 

 

Tries to get `y.index` (works if this is a pd.Series), if that 

fails, return np.arange(y.shape[0]). 

 

This will be extended in the future to deal with more types of 

labeled data. 

 

Parameters 

---------- 

y : scalar or array-like 

The proposed y-value 

 

Returns 

------- 

x, y : ndarray 

The x and y values to plot. 

""" 

try: 

return y.index.values, y.values 

except AttributeError: 

y = _check_1d(y) 

return np.arange(y.shape[0], dtype=float), y 

 

 

def safe_first_element(obj): 

if isinstance(obj, collections.Iterator): 

# needed to accept `array.flat` as input. 

# np.flatiter reports as an instance of collections.Iterator 

# but can still be indexed via []. 

# This has the side effect of re-setting the iterator, but 

# that is acceptable. 

try: 

return obj[0] 

except TypeError: 

pass 

raise RuntimeError("matplotlib does not support generators " 

"as input") 

return next(iter(obj)) 

 

 

def sanitize_sequence(data): 

"""Converts dictview object to list""" 

return list(data) if isinstance(data, collections.MappingView) else data 

 

 

def normalize_kwargs(kw, alias_mapping=None, required=(), forbidden=(), 

allowed=None): 

"""Helper function to normalize kwarg inputs 

 

The order they are resolved are: 

 

1. aliasing 

2. required 

3. forbidden 

4. allowed 

 

This order means that only the canonical names need appear in 

`allowed`, `forbidden`, `required` 

 

Parameters 

---------- 

 

alias_mapping, dict, optional 

A mapping between a canonical name to a list of 

aliases, in order of precedence from lowest to highest. 

 

If the canonical value is not in the list it is assumed to have 

the highest priority. 

 

required : iterable, optional 

A tuple of fields that must be in kwargs. 

 

forbidden : iterable, optional 

A list of keys which may not be in kwargs 

 

allowed : tuple, optional 

A tuple of allowed fields. If this not None, then raise if 

`kw` contains any keys not in the union of `required` 

and `allowed`. To allow only the required fields pass in 

``()`` for `allowed` 

 

Raises 

------ 

TypeError 

To match what python raises if invalid args/kwargs are passed to 

a callable. 

 

""" 

# deal with default value of alias_mapping 

if alias_mapping is None: 

alias_mapping = dict() 

 

# make a local so we can pop 

kw = dict(kw) 

# output dictionary 

ret = dict() 

 

# hit all alias mappings 

for canonical, alias_list in six.iteritems(alias_mapping): 

 

# the alias lists are ordered from lowest to highest priority 

# so we know to use the last value in this list 

tmp = [] 

seen = [] 

for a in alias_list: 

try: 

tmp.append(kw.pop(a)) 

seen.append(a) 

except KeyError: 

pass 

# if canonical is not in the alias_list assume highest priority 

if canonical not in alias_list: 

try: 

tmp.append(kw.pop(canonical)) 

seen.append(canonical) 

except KeyError: 

pass 

# if we found anything in this set of aliases put it in the return 

# dict 

if tmp: 

ret[canonical] = tmp[-1] 

if len(tmp) > 1: 

warnings.warn("Saw kwargs {seen!r} which are all aliases for " 

"{canon!r}. Kept value from {used!r}".format( 

seen=seen, canon=canonical, used=seen[-1])) 

 

# at this point we know that all keys which are aliased are removed, update 

# the return dictionary from the cleaned local copy of the input 

ret.update(kw) 

 

fail_keys = [k for k in required if k not in ret] 

if fail_keys: 

raise TypeError("The required keys {keys!r} " 

"are not in kwargs".format(keys=fail_keys)) 

 

fail_keys = [k for k in forbidden if k in ret] 

if fail_keys: 

raise TypeError("The forbidden keys {keys!r} " 

"are in kwargs".format(keys=fail_keys)) 

 

if allowed is not None: 

allowed_set = set(required) | set(allowed) 

fail_keys = [k for k in ret if k not in allowed_set] 

if fail_keys: 

raise TypeError("kwargs contains {keys!r} which are not in " 

"the required {req!r} or " 

"allowed {allow!r} keys".format( 

keys=fail_keys, req=required, 

allow=allowed)) 

 

return ret 

 

 

def get_label(y, default_name): 

try: 

return y.name 

except AttributeError: 

return default_name 

 

 

_lockstr = """\ 

LOCKERROR: matplotlib is trying to acquire the lock 

{!r} 

and has failed. This maybe due to any other process holding this 

lock. If you are sure no other matplotlib process is running try 

removing these folders and trying again. 

""" 

 

 

class Locked(object): 

""" 

Context manager to handle locks. 

 

Based on code from conda. 

 

(c) 2012-2013 Continuum Analytics, Inc. / https://www.continuum.io/ 

All Rights Reserved 

 

conda is distributed under the terms of the BSD 3-clause license. 

Consult LICENSE_CONDA or https://opensource.org/licenses/BSD-3-Clause. 

""" 

LOCKFN = '.matplotlib_lock' 

 

class TimeoutError(RuntimeError): 

pass 

 

def __init__(self, path): 

self.path = path 

self.end = "-" + str(os.getpid()) 

self.lock_path = os.path.join(self.path, self.LOCKFN + self.end) 

self.pattern = os.path.join(self.path, self.LOCKFN + '-*') 

self.remove = True 

 

def __enter__(self): 

retries = 50 

sleeptime = 0.1 

while retries: 

files = glob.glob(self.pattern) 

if files and not files[0].endswith(self.end): 

time.sleep(sleeptime) 

retries -= 1 

else: 

break 

else: 

err_str = _lockstr.format(self.pattern) 

raise self.TimeoutError(err_str) 

 

if not files: 

try: 

os.makedirs(self.lock_path) 

except OSError: 

pass 

else: # PID lock already here --- someone else will remove it. 

self.remove = False 

 

def __exit__(self, exc_type, exc_value, traceback): 

if self.remove: 

for path in self.lock_path, self.path: 

try: 

os.rmdir(path) 

except OSError: 

pass 

 

 

class _FuncInfo(object): 

""" 

Class used to store a function. 

 

""" 

 

def __init__(self, function, inverse, bounded_0_1=True, check_params=None): 

""" 

Parameters 

---------- 

 

function : callable 

A callable implementing the function receiving the variable as 

first argument and any additional parameters in a list as second 

argument. 

inverse : callable 

A callable implementing the inverse function receiving the variable 

as first argument and any additional parameters in a list as 

second argument. It must satisfy 'inverse(function(x, p), p) == x'. 

bounded_0_1: bool or callable 

A boolean indicating whether the function is bounded in the [0,1] 

interval, or a callable taking a list of values for the additional 

parameters, and returning a boolean indicating whether the function 

is bounded in the [0,1] interval for that combination of 

parameters. Default True. 

check_params: callable or None 

A callable taking a list of values for the additional parameters 

and returning a boolean indicating whether that combination of 

parameters is valid. It is only required if the function has 

additional parameters and some of them are restricted. 

Default None. 

 

""" 

 

self.function = function 

self.inverse = inverse 

 

if callable(bounded_0_1): 

self._bounded_0_1 = bounded_0_1 

else: 

self._bounded_0_1 = lambda x: bounded_0_1 

 

if check_params is None: 

self._check_params = lambda x: True 

elif callable(check_params): 

self._check_params = check_params 

else: 

raise ValueError("Invalid 'check_params' argument.") 

 

def is_bounded_0_1(self, params=None): 

""" 

Returns a boolean indicating if the function is bounded in the [0,1] 

interval for a particular set of additional parameters. 

 

Parameters 

---------- 

 

params : list 

The list of additional parameters. Default None. 

 

Returns 

------- 

 

out : bool 

True if the function is bounded in the [0,1] interval for 

parameters 'params'. Otherwise False. 

 

""" 

 

return self._bounded_0_1(params) 

 

def check_params(self, params=None): 

""" 

Returns a boolean indicating if the set of additional parameters is 

valid. 

 

Parameters 

---------- 

 

params : list 

The list of additional parameters. Default None. 

 

Returns 

------- 

 

out : bool 

True if 'params' is a valid set of additional parameters for the 

function. Otherwise False. 

 

""" 

 

return self._check_params(params) 

 

 

class _StringFuncParser(object): 

""" 

A class used to convert predefined strings into 

_FuncInfo objects, or to directly obtain _FuncInfo 

properties. 

 

""" 

 

_funcs = {} 

_funcs['linear'] = _FuncInfo(lambda x: x, 

lambda x: x, 

True) 

_funcs['quadratic'] = _FuncInfo(np.square, 

np.sqrt, 

True) 

_funcs['cubic'] = _FuncInfo(lambda x: x**3, 

lambda x: x**(1. / 3), 

True) 

_funcs['sqrt'] = _FuncInfo(np.sqrt, 

np.square, 

True) 

_funcs['cbrt'] = _FuncInfo(lambda x: x**(1. / 3), 

lambda x: x**3, 

True) 

_funcs['log10'] = _FuncInfo(np.log10, 

lambda x: (10**(x)), 

False) 

_funcs['log'] = _FuncInfo(np.log, 

np.exp, 

False) 

_funcs['log2'] = _FuncInfo(np.log2, 

lambda x: (2**x), 

False) 

_funcs['x**{p}'] = _FuncInfo(lambda x, p: x**p[0], 

lambda x, p: x**(1. / p[0]), 

True) 

_funcs['root{p}(x)'] = _FuncInfo(lambda x, p: x**(1. / p[0]), 

lambda x, p: x**p, 

True) 

_funcs['log{p}(x)'] = _FuncInfo(lambda x, p: (np.log(x) / 

np.log(p[0])), 

lambda x, p: p[0]**(x), 

False, 

lambda p: p[0] > 0) 

_funcs['log10(x+{p})'] = _FuncInfo(lambda x, p: np.log10(x + p[0]), 

lambda x, p: 10**x - p[0], 

lambda p: p[0] > 0) 

_funcs['log(x+{p})'] = _FuncInfo(lambda x, p: np.log(x + p[0]), 

lambda x, p: np.exp(x) - p[0], 

lambda p: p[0] > 0) 

_funcs['log{p}(x+{p})'] = _FuncInfo(lambda x, p: (np.log(x + p[1]) / 

np.log(p[0])), 

lambda x, p: p[0]**(x) - p[1], 

lambda p: p[1] > 0, 

lambda p: p[0] > 0) 

 

def __init__(self, str_func): 

""" 

Parameters 

---------- 

str_func : string 

String to be parsed. 

 

""" 

 

if not isinstance(str_func, six.string_types): 

raise ValueError("'%s' must be a string." % str_func) 

self._str_func = six.text_type(str_func) 

self._key, self._params = self._get_key_params() 

self._func = self._parse_func() 

 

def _parse_func(self): 

""" 

Parses the parameters to build a new _FuncInfo object, 

replacing the relevant parameters if necessary in the lambda 

functions. 

 

""" 

 

func = self._funcs[self._key] 

 

if not self._params: 

func = _FuncInfo(func.function, func.inverse, 

func.is_bounded_0_1()) 

else: 

m = func.function 

function = (lambda x, m=m: m(x, self._params)) 

 

m = func.inverse 

inverse = (lambda x, m=m: m(x, self._params)) 

 

is_bounded_0_1 = func.is_bounded_0_1(self._params) 

 

func = _FuncInfo(function, inverse, 

is_bounded_0_1) 

return func 

 

@property 

def func_info(self): 

""" 

Returns the _FuncInfo object. 

 

""" 

return self._func 

 

@property 

def function(self): 

""" 

Returns the callable for the direct function. 

 

""" 

return self._func.function 

 

@property 

def inverse(self): 

""" 

Returns the callable for the inverse function. 

 

""" 

return self._func.inverse 

 

@property 

def is_bounded_0_1(self): 

""" 

Returns a boolean indicating if the function is bounded 

in the [0-1 interval]. 

 

""" 

return self._func.is_bounded_0_1() 

 

def _get_key_params(self): 

str_func = self._str_func 

# Checking if it comes with parameters 

regex = r'\{(.*?)\}' 

params = re.findall(regex, str_func) 

 

for i, param in enumerate(params): 

try: 

params[i] = float(param) 

except ValueError: 

raise ValueError("Parameter %i is '%s', which is " 

"not a number." % 

(i, param)) 

 

str_func = re.sub(regex, '{p}', str_func) 

 

try: 

func = self._funcs[str_func] 

except (ValueError, KeyError): 

raise ValueError("'%s' is an invalid string. The only strings " 

"recognized as functions are %s." % 

(str_func, list(self._funcs))) 

 

# Checking that the parameters are valid 

if not func.check_params(params): 

raise ValueError("%s are invalid values for the parameters " 

"in %s." % 

(params, str_func)) 

 

return str_func, params 

 

 

def _topmost_artist( 

artists, 

_cached_max=functools.partial(max, key=operator.attrgetter("zorder"))): 

"""Get the topmost artist of a list. 

 

In case of a tie, return the *last* of the tied artists, as it will be 

drawn on top of the others. `max` returns the first maximum in case of ties 

(on Py2 this is undocumented but true), so we need to iterate over the list 

in reverse order. 

""" 

return _cached_max(reversed(artists)) 

 

 

def _str_equal(obj, s): 

"""Return whether *obj* is a string equal to string *s*. 

 

This helper solely exists to handle the case where *obj* is a numpy array, 

because in such cases, a naive ``obj == s`` would yield an array, which 

cannot be used in a boolean context. 

""" 

return isinstance(obj, six.string_types) and obj == s 

 

 

def _str_lower_equal(obj, s): 

"""Return whether *obj* is a string equal, when lowercased, to string *s*. 

 

This helper solely exists to handle the case where *obj* is a numpy array, 

because in such cases, a naive ``obj == s`` would yield an array, which 

cannot be used in a boolean context. 

""" 

return isinstance(obj, six.string_types) and obj.lower() == s