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

""" 

Discrete Fourier Transforms - basic.py 

""" 

# Created by Pearu Peterson, August,September 2002 

from __future__ import division, print_function, absolute_import 

 

__all__ = ['fft','ifft','fftn','ifftn','rfft','irfft', 

'fft2','ifft2'] 

 

from numpy import zeros, swapaxes 

import numpy 

from . import _fftpack 

 

import atexit 

atexit.register(_fftpack.destroy_zfft_cache) 

atexit.register(_fftpack.destroy_zfftnd_cache) 

atexit.register(_fftpack.destroy_drfft_cache) 

atexit.register(_fftpack.destroy_cfft_cache) 

atexit.register(_fftpack.destroy_cfftnd_cache) 

atexit.register(_fftpack.destroy_rfft_cache) 

del atexit 

 

 

def istype(arr, typeclass): 

return issubclass(arr.dtype.type, typeclass) 

 

 

def _datacopied(arr, original): 

""" 

Strict check for `arr` not sharing any data with `original`, 

under the assumption that arr = asarray(original) 

 

""" 

if arr is original: 

return False 

if not isinstance(original, numpy.ndarray) and hasattr(original, '__array__'): 

return False 

return arr.base is None 

 

# XXX: single precision FFTs partially disabled due to accuracy issues 

# for large prime-sized inputs. 

# 

# See http://permalink.gmane.org/gmane.comp.python.scientific.devel/13834 

# ("fftpack test failures for 0.8.0b1", Ralf Gommers, 17 Jun 2010, 

# @ scipy-dev) 

# 

# These should be re-enabled once the problems are resolved 

 

 

def _is_safe_size(n): 

""" 

Is the size of FFT such that FFTPACK can handle it in single precision 

with sufficient accuracy? 

 

Composite numbers of 2, 3, and 5 are accepted, as FFTPACK has those 

""" 

n = int(n) 

 

if n == 0: 

return True 

 

# Divide by 3 until you can't, then by 5 until you can't 

for c in (3, 5): 

while n % c == 0: 

n //= c 

 

# Return True if the remainder is a power of 2 

return not n & (n-1) 

 

 

def _fake_crfft(x, n, *a, **kw): 

if _is_safe_size(n): 

return _fftpack.crfft(x, n, *a, **kw) 

else: 

return _fftpack.zrfft(x, n, *a, **kw).astype(numpy.complex64) 

 

 

def _fake_cfft(x, n, *a, **kw): 

if _is_safe_size(n): 

return _fftpack.cfft(x, n, *a, **kw) 

else: 

return _fftpack.zfft(x, n, *a, **kw).astype(numpy.complex64) 

 

 

def _fake_rfft(x, n, *a, **kw): 

if _is_safe_size(n): 

return _fftpack.rfft(x, n, *a, **kw) 

else: 

return _fftpack.drfft(x, n, *a, **kw).astype(numpy.float32) 

 

 

def _fake_cfftnd(x, shape, *a, **kw): 

if numpy.all(list(map(_is_safe_size, shape))): 

return _fftpack.cfftnd(x, shape, *a, **kw) 

else: 

return _fftpack.zfftnd(x, shape, *a, **kw).astype(numpy.complex64) 

 

 

_DTYPE_TO_FFT = { 

# numpy.dtype(numpy.float32): _fftpack.crfft, 

numpy.dtype(numpy.float32): _fake_crfft, 

numpy.dtype(numpy.float64): _fftpack.zrfft, 

# numpy.dtype(numpy.complex64): _fftpack.cfft, 

numpy.dtype(numpy.complex64): _fake_cfft, 

numpy.dtype(numpy.complex128): _fftpack.zfft, 

} 

 

_DTYPE_TO_RFFT = { 

# numpy.dtype(numpy.float32): _fftpack.rfft, 

numpy.dtype(numpy.float32): _fake_rfft, 

numpy.dtype(numpy.float64): _fftpack.drfft, 

} 

 

_DTYPE_TO_FFTN = { 

# numpy.dtype(numpy.complex64): _fftpack.cfftnd, 

numpy.dtype(numpy.complex64): _fake_cfftnd, 

numpy.dtype(numpy.complex128): _fftpack.zfftnd, 

# numpy.dtype(numpy.float32): _fftpack.cfftnd, 

numpy.dtype(numpy.float32): _fake_cfftnd, 

numpy.dtype(numpy.float64): _fftpack.zfftnd, 

} 

 

 

def _asfarray(x): 

"""Like numpy asfarray, except that it does not modify x dtype if x is 

already an array with a float dtype, and do not cast complex types to 

real.""" 

if hasattr(x, "dtype") and x.dtype.char in numpy.typecodes["AllFloat"]: 

# 'dtype' attribute does not ensure that the 

# object is an ndarray (e.g. Series class 

# from the pandas library) 

if x.dtype == numpy.half: 

# no half-precision routines, so convert to single precision 

return numpy.asarray(x, dtype=numpy.float32) 

return numpy.asarray(x, dtype=x.dtype) 

else: 

# We cannot use asfarray directly because it converts sequences of 

# complex to sequence of real 

ret = numpy.asarray(x) 

if ret.dtype == numpy.half: 

return numpy.asarray(ret, dtype=numpy.float32) 

elif ret.dtype.char not in numpy.typecodes["AllFloat"]: 

return numpy.asfarray(x) 

return ret 

 

 

def _fix_shape(x, n, axis): 

""" Internal auxiliary function for _raw_fft, _raw_fftnd.""" 

s = list(x.shape) 

if s[axis] > n: 

index = [slice(None)]*len(s) 

index[axis] = slice(0,n) 

x = x[index] 

return x, False 

else: 

index = [slice(None)]*len(s) 

index[axis] = slice(0,s[axis]) 

s[axis] = n 

z = zeros(s,x.dtype.char) 

z[index] = x 

return z, True 

 

 

def _raw_fft(x, n, axis, direction, overwrite_x, work_function): 

""" Internal auxiliary function for fft, ifft, rfft, irfft.""" 

if n is None: 

n = x.shape[axis] 

elif n != x.shape[axis]: 

x, copy_made = _fix_shape(x,n,axis) 

overwrite_x = overwrite_x or copy_made 

 

if n < 1: 

raise ValueError("Invalid number of FFT data points " 

"(%d) specified." % n) 

 

if axis == -1 or axis == len(x.shape)-1: 

r = work_function(x,n,direction,overwrite_x=overwrite_x) 

else: 

x = swapaxes(x, axis, -1) 

r = work_function(x,n,direction,overwrite_x=overwrite_x) 

r = swapaxes(r, axis, -1) 

return r 

 

 

def fft(x, n=None, axis=-1, overwrite_x=False): 

""" 

Return discrete Fourier transform of real or complex sequence. 

 

The returned complex array contains ``y(0), y(1),..., y(n-1)`` where 

 

``y(j) = (x * exp(-2*pi*sqrt(-1)*j*np.arange(n)/n)).sum()``. 

 

Parameters 

---------- 

x : array_like 

Array to Fourier transform. 

n : int, optional 

Length of the Fourier transform. If ``n < x.shape[axis]``, `x` is 

truncated. If ``n > x.shape[axis]``, `x` is zero-padded. The 

default results in ``n = x.shape[axis]``. 

axis : int, optional 

Axis along which the fft's are computed; the default is over the 

last axis (i.e., ``axis=-1``). 

overwrite_x : bool, optional 

If True, the contents of `x` can be destroyed; the default is False. 

 

Returns 

------- 

z : complex ndarray 

with the elements:: 

 

[y(0),y(1),..,y(n/2),y(1-n/2),...,y(-1)] if n is even 

[y(0),y(1),..,y((n-1)/2),y(-(n-1)/2),...,y(-1)] if n is odd 

 

where:: 

 

y(j) = sum[k=0..n-1] x[k] * exp(-sqrt(-1)*j*k* 2*pi/n), j = 0..n-1 

 

See Also 

-------- 

ifft : Inverse FFT 

rfft : FFT of a real sequence 

 

Notes 

----- 

The packing of the result is "standard": If ``A = fft(a, n)``, then 

``A[0]`` contains the zero-frequency term, ``A[1:n/2]`` contains the 

positive-frequency terms, and ``A[n/2:]`` contains the negative-frequency 

terms, in order of decreasingly negative frequency. So for an 8-point 

transform, the frequencies of the result are [0, 1, 2, 3, -4, -3, -2, -1]. 

To rearrange the fft output so that the zero-frequency component is 

centered, like [-4, -3, -2, -1, 0, 1, 2, 3], use `fftshift`. 

 

Both single and double precision routines are implemented. Half precision 

inputs will be converted to single precision. Non floating-point inputs 

will be converted to double precision. Long-double precision inputs are 

not supported. 

 

This function is most efficient when `n` is a power of two, and least 

efficient when `n` is prime. 

 

Note that if ``x`` is real-valued then ``A[j] == A[n-j].conjugate()``. 

If ``x`` is real-valued and ``n`` is even then ``A[n/2]`` is real. 

 

If the data type of `x` is real, a "real FFT" algorithm is automatically 

used, which roughly halves the computation time. To increase efficiency 

a little further, use `rfft`, which does the same calculation, but only 

outputs half of the symmetrical spectrum. If the data is both real and 

symmetrical, the `dct` can again double the efficiency, by generating 

half of the spectrum from half of the signal. 

 

Examples 

-------- 

>>> from scipy.fftpack import fft, ifft 

>>> x = np.arange(5) 

>>> np.allclose(fft(ifft(x)), x, atol=1e-15) # within numerical accuracy. 

True 

 

""" 

tmp = _asfarray(x) 

 

try: 

work_function = _DTYPE_TO_FFT[tmp.dtype] 

except KeyError: 

raise ValueError("type %s is not supported" % tmp.dtype) 

 

if not (istype(tmp, numpy.complex64) or istype(tmp, numpy.complex128)): 

overwrite_x = 1 

 

overwrite_x = overwrite_x or _datacopied(tmp, x) 

 

if n is None: 

n = tmp.shape[axis] 

elif n != tmp.shape[axis]: 

tmp, copy_made = _fix_shape(tmp,n,axis) 

overwrite_x = overwrite_x or copy_made 

 

if n < 1: 

raise ValueError("Invalid number of FFT data points " 

"(%d) specified." % n) 

 

if axis == -1 or axis == len(tmp.shape) - 1: 

return work_function(tmp,n,1,0,overwrite_x) 

 

tmp = swapaxes(tmp, axis, -1) 

tmp = work_function(tmp,n,1,0,overwrite_x) 

return swapaxes(tmp, axis, -1) 

 

 

def ifft(x, n=None, axis=-1, overwrite_x=False): 

""" 

Return discrete inverse Fourier transform of real or complex sequence. 

 

The returned complex array contains ``y(0), y(1),..., y(n-1)`` where 

 

``y(j) = (x * exp(2*pi*sqrt(-1)*j*np.arange(n)/n)).mean()``. 

 

Parameters 

---------- 

x : array_like 

Transformed data to invert. 

n : int, optional 

Length of the inverse Fourier transform. If ``n < x.shape[axis]``, 

`x` is truncated. If ``n > x.shape[axis]``, `x` is zero-padded. 

The default results in ``n = x.shape[axis]``. 

axis : int, optional 

Axis along which the ifft's are computed; the default is over the 

last axis (i.e., ``axis=-1``). 

overwrite_x : bool, optional 

If True, the contents of `x` can be destroyed; the default is False. 

 

Returns 

------- 

ifft : ndarray of floats 

The inverse discrete Fourier transform. 

 

See Also 

-------- 

fft : Forward FFT 

 

Notes 

----- 

Both single and double precision routines are implemented. Half precision 

inputs will be converted to single precision. Non floating-point inputs 

will be converted to double precision. Long-double precision inputs are 

not supported. 

 

This function is most efficient when `n` is a power of two, and least 

efficient when `n` is prime. 

 

If the data type of `x` is real, a "real IFFT" algorithm is automatically 

used, which roughly halves the computation time. 

 

Examples 

-------- 

>>> from scipy.fftpack import fft, ifft 

>>> import numpy as np 

>>> x = np.arange(5) 

>>> np.allclose(ifft(fft(x)), x, atol=1e-15) # within numerical accuracy. 

True 

 

""" 

tmp = _asfarray(x) 

 

try: 

work_function = _DTYPE_TO_FFT[tmp.dtype] 

except KeyError: 

raise ValueError("type %s is not supported" % tmp.dtype) 

 

if not (istype(tmp, numpy.complex64) or istype(tmp, numpy.complex128)): 

overwrite_x = 1 

 

overwrite_x = overwrite_x or _datacopied(tmp, x) 

 

if n is None: 

n = tmp.shape[axis] 

elif n != tmp.shape[axis]: 

tmp, copy_made = _fix_shape(tmp,n,axis) 

overwrite_x = overwrite_x or copy_made 

 

if n < 1: 

raise ValueError("Invalid number of FFT data points " 

"(%d) specified." % n) 

 

if axis == -1 or axis == len(tmp.shape) - 1: 

return work_function(tmp,n,-1,1,overwrite_x) 

 

tmp = swapaxes(tmp, axis, -1) 

tmp = work_function(tmp,n,-1,1,overwrite_x) 

return swapaxes(tmp, axis, -1) 

 

 

def rfft(x, n=None, axis=-1, overwrite_x=False): 

""" 

Discrete Fourier transform of a real sequence. 

 

Parameters 

---------- 

x : array_like, real-valued 

The data to transform. 

n : int, optional 

Defines the length of the Fourier transform. If `n` is not specified 

(the default) then ``n = x.shape[axis]``. If ``n < x.shape[axis]``, 

`x` is truncated, if ``n > x.shape[axis]``, `x` is zero-padded. 

axis : int, optional 

The axis along which the transform is applied. The default is the 

last axis. 

overwrite_x : bool, optional 

If set to true, the contents of `x` can be overwritten. Default is 

False. 

 

Returns 

------- 

z : real ndarray 

The returned real array contains:: 

 

[y(0),Re(y(1)),Im(y(1)),...,Re(y(n/2))] if n is even 

[y(0),Re(y(1)),Im(y(1)),...,Re(y(n/2)),Im(y(n/2))] if n is odd 

 

where:: 

 

y(j) = sum[k=0..n-1] x[k] * exp(-sqrt(-1)*j*k*2*pi/n) 

j = 0..n-1 

 

See Also 

-------- 

fft, irfft, numpy.fft.rfft 

 

Notes 

----- 

Within numerical accuracy, ``y == rfft(irfft(y))``. 

 

Both single and double precision routines are implemented. Half precision 

inputs will be converted to single precision. Non floating-point inputs 

will be converted to double precision. Long-double precision inputs are 

not supported. 

 

To get an output with a complex datatype, consider using the related 

function `numpy.fft.rfft`. 

 

Examples 

-------- 

>>> from scipy.fftpack import fft, rfft 

>>> a = [9, -9, 1, 3] 

>>> fft(a) 

array([ 4. +0.j, 8.+12.j, 16. +0.j, 8.-12.j]) 

>>> rfft(a) 

array([ 4., 8., 12., 16.]) 

 

""" 

tmp = _asfarray(x) 

 

if not numpy.isrealobj(tmp): 

raise TypeError("1st argument must be real sequence") 

 

try: 

work_function = _DTYPE_TO_RFFT[tmp.dtype] 

except KeyError: 

raise ValueError("type %s is not supported" % tmp.dtype) 

 

overwrite_x = overwrite_x or _datacopied(tmp, x) 

 

return _raw_fft(tmp,n,axis,1,overwrite_x,work_function) 

 

 

def irfft(x, n=None, axis=-1, overwrite_x=False): 

""" 

Return inverse discrete Fourier transform of real sequence x. 

 

The contents of `x` are interpreted as the output of the `rfft` 

function. 

 

Parameters 

---------- 

x : array_like 

Transformed data to invert. 

n : int, optional 

Length of the inverse Fourier transform. 

If n < x.shape[axis], x is truncated. 

If n > x.shape[axis], x is zero-padded. 

The default results in n = x.shape[axis]. 

axis : int, optional 

Axis along which the ifft's are computed; the default is over 

the last axis (i.e., axis=-1). 

overwrite_x : bool, optional 

If True, the contents of `x` can be destroyed; the default is False. 

 

Returns 

------- 

irfft : ndarray of floats 

The inverse discrete Fourier transform. 

 

See Also 

-------- 

rfft, ifft, numpy.fft.irfft 

 

Notes 

----- 

The returned real array contains:: 

 

[y(0),y(1),...,y(n-1)] 

 

where for n is even:: 

 

y(j) = 1/n (sum[k=1..n/2-1] (x[2*k-1]+sqrt(-1)*x[2*k]) 

* exp(sqrt(-1)*j*k* 2*pi/n) 

+ c.c. + x[0] + (-1)**(j) x[n-1]) 

 

and for n is odd:: 

 

y(j) = 1/n (sum[k=1..(n-1)/2] (x[2*k-1]+sqrt(-1)*x[2*k]) 

* exp(sqrt(-1)*j*k* 2*pi/n) 

+ c.c. + x[0]) 

 

c.c. denotes complex conjugate of preceding expression. 

 

For details on input parameters, see `rfft`. 

 

To process (conjugate-symmetric) frequency-domain data with a complex 

datatype, consider using the related function `numpy.fft.irfft`. 

""" 

tmp = _asfarray(x) 

if not numpy.isrealobj(tmp): 

raise TypeError("1st argument must be real sequence") 

 

try: 

work_function = _DTYPE_TO_RFFT[tmp.dtype] 

except KeyError: 

raise ValueError("type %s is not supported" % tmp.dtype) 

 

overwrite_x = overwrite_x or _datacopied(tmp, x) 

 

return _raw_fft(tmp,n,axis,-1,overwrite_x,work_function) 

 

 

def _raw_fftnd(x, s, axes, direction, overwrite_x, work_function): 

""" Internal auxiliary function for fftnd, ifftnd.""" 

if s is None: 

if axes is None: 

s = x.shape 

else: 

s = numpy.take(x.shape, axes) 

 

s = tuple(s) 

if axes is None: 

noaxes = True 

axes = list(range(-x.ndim, 0)) 

else: 

noaxes = False 

if len(axes) != len(s): 

raise ValueError("when given, axes and shape arguments " 

"have to be of the same length") 

 

for dim in s: 

if dim < 1: 

raise ValueError("Invalid number of FFT data points " 

"(%s) specified." % (s,)) 

 

# No need to swap axes, array is in C order 

if noaxes: 

for i in axes: 

x, copy_made = _fix_shape(x, s[i], i) 

overwrite_x = overwrite_x or copy_made 

return work_function(x,s,direction,overwrite_x=overwrite_x) 

 

# We ordered axes, because the code below to push axes at the end of the 

# array assumes axes argument is in ascending order. 

a = numpy.array(axes, numpy.intc) 

abs_axes = numpy.where(a < 0, a + x.ndim, a) 

id_ = numpy.argsort(abs_axes) 

axes = [axes[i] for i in id_] 

s = [s[i] for i in id_] 

 

# Swap the request axes, last first (i.e. First swap the axis which ends up 

# at -1, then at -2, etc...), such as the request axes on which the 

# operation is carried become the last ones 

for i in range(1, len(axes)+1): 

x = numpy.swapaxes(x, axes[-i], -i) 

 

# We can now operate on the axes waxes, the p last axes (p = len(axes)), by 

# fixing the shape of the input array to 1 for any axis the fft is not 

# carried upon. 

waxes = list(range(x.ndim - len(axes), x.ndim)) 

shape = numpy.ones(x.ndim) 

shape[waxes] = s 

 

for i in range(len(waxes)): 

x, copy_made = _fix_shape(x, s[i], waxes[i]) 

overwrite_x = overwrite_x or copy_made 

 

r = work_function(x, shape, direction, overwrite_x=overwrite_x) 

 

# reswap in the reverse order (first axis first, etc...) to get original 

# order 

for i in range(len(axes), 0, -1): 

r = numpy.swapaxes(r, -i, axes[-i]) 

 

return r 

 

 

def fftn(x, shape=None, axes=None, overwrite_x=False): 

""" 

Return multidimensional discrete Fourier transform. 

 

The returned array contains:: 

 

y[j_1,..,j_d] = sum[k_1=0..n_1-1, ..., k_d=0..n_d-1] 

x[k_1,..,k_d] * prod[i=1..d] exp(-sqrt(-1)*2*pi/n_i * j_i * k_i) 

 

where d = len(x.shape) and n = x.shape. 

 

Parameters 

---------- 

x : array_like 

The (n-dimensional) array to transform. 

shape : tuple of ints, optional 

The shape of the result. If both `shape` and `axes` (see below) are 

None, `shape` is ``x.shape``; if `shape` is None but `axes` is 

not None, then `shape` is ``scipy.take(x.shape, axes, axis=0)``. 

If ``shape[i] > x.shape[i]``, the i-th dimension is padded with zeros. 

If ``shape[i] < x.shape[i]``, the i-th dimension is truncated to 

length ``shape[i]``. 

axes : array_like of ints, optional 

The axes of `x` (`y` if `shape` is not None) along which the 

transform is applied. 

overwrite_x : bool, optional 

If True, the contents of `x` can be destroyed. Default is False. 

 

Returns 

------- 

y : complex-valued n-dimensional numpy array 

The (n-dimensional) DFT of the input array. 

 

See Also 

-------- 

ifftn 

 

Notes 

----- 

If ``x`` is real-valued, then 

``y[..., j_i, ...] == y[..., n_i-j_i, ...].conjugate()``. 

 

Both single and double precision routines are implemented. Half precision 

inputs will be converted to single precision. Non floating-point inputs 

will be converted to double precision. Long-double precision inputs are 

not supported. 

 

Examples 

-------- 

>>> from scipy.fftpack import fftn, ifftn 

>>> y = (-np.arange(16), 8 - np.arange(16), np.arange(16)) 

>>> np.allclose(y, fftn(ifftn(y))) 

True 

 

""" 

return _raw_fftn_dispatch(x, shape, axes, overwrite_x, 1) 

 

 

def _raw_fftn_dispatch(x, shape, axes, overwrite_x, direction): 

tmp = _asfarray(x) 

 

try: 

work_function = _DTYPE_TO_FFTN[tmp.dtype] 

except KeyError: 

raise ValueError("type %s is not supported" % tmp.dtype) 

 

if not (istype(tmp, numpy.complex64) or istype(tmp, numpy.complex128)): 

overwrite_x = 1 

 

overwrite_x = overwrite_x or _datacopied(tmp, x) 

return _raw_fftnd(tmp,shape,axes,direction,overwrite_x,work_function) 

 

 

def ifftn(x, shape=None, axes=None, overwrite_x=False): 

""" 

Return inverse multi-dimensional discrete Fourier transform of 

arbitrary type sequence x. 

 

The returned array contains:: 

 

y[j_1,..,j_d] = 1/p * sum[k_1=0..n_1-1, ..., k_d=0..n_d-1] 

x[k_1,..,k_d] * prod[i=1..d] exp(sqrt(-1)*2*pi/n_i * j_i * k_i) 

 

where ``d = len(x.shape)``, ``n = x.shape``, and ``p = prod[i=1..d] n_i``. 

 

For description of parameters see `fftn`. 

 

See Also 

-------- 

fftn : for detailed information. 

 

Examples 

-------- 

>>> from scipy.fftpack import fftn, ifftn 

>>> import numpy as np 

>>> y = (-np.arange(16), 8 - np.arange(16), np.arange(16)) 

>>> np.allclose(y, ifftn(fftn(y))) 

True 

 

""" 

return _raw_fftn_dispatch(x, shape, axes, overwrite_x, -1) 

 

 

def fft2(x, shape=None, axes=(-2,-1), overwrite_x=False): 

""" 

2-D discrete Fourier transform. 

 

Return the two-dimensional discrete Fourier transform of the 2-D argument 

`x`. 

 

See Also 

-------- 

fftn : for detailed information. 

 

""" 

return fftn(x,shape,axes,overwrite_x) 

 

 

def ifft2(x, shape=None, axes=(-2,-1), overwrite_x=False): 

""" 

2-D discrete inverse Fourier transform of real or complex sequence. 

 

Return inverse two-dimensional discrete Fourier transform of 

arbitrary type sequence x. 

 

See `ifft` for more information. 

 

See also 

-------- 

fft2, ifft 

 

""" 

return ifftn(x,shape,axes,overwrite_x)