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

from __future__ import division, print_function, absolute_import 

import inspect 

import numpy as np 

from .bdf import BDF 

from .radau import Radau 

from .rk import RK23, RK45 

from .lsoda import LSODA 

from scipy.optimize import OptimizeResult 

from .common import EPS, OdeSolution 

from .base import OdeSolver 

 

 

METHODS = {'RK23': RK23, 

'RK45': RK45, 

'Radau': Radau, 

'BDF': BDF, 

'LSODA': LSODA} 

 

 

MESSAGES = {0: "The solver successfully reached the end of the integration interval.", 

1: "A termination event occurred."} 

 

 

class OdeResult(OptimizeResult): 

pass 

 

 

def prepare_events(events): 

"""Standardize event functions and extract is_terminal and direction.""" 

if callable(events): 

events = (events,) 

 

if events is not None: 

is_terminal = np.empty(len(events), dtype=bool) 

direction = np.empty(len(events)) 

for i, event in enumerate(events): 

try: 

is_terminal[i] = event.terminal 

except AttributeError: 

is_terminal[i] = False 

 

try: 

direction[i] = event.direction 

except AttributeError: 

direction[i] = 0 

else: 

is_terminal = None 

direction = None 

 

return events, is_terminal, direction 

 

 

def solve_event_equation(event, sol, t_old, t): 

"""Solve an equation corresponding to an ODE event. 

 

The equation is ``event(t, y(t)) = 0``, here ``y(t)`` is known from an 

ODE solver using some sort of interpolation. It is solved by 

`scipy.optimize.brentq` with xtol=atol=4*EPS. 

 

Parameters 

---------- 

event : callable 

Function ``event(t, y)``. 

sol : callable 

Function ``sol(t)`` which evaluates an ODE solution between `t_old` 

and `t`. 

t_old, t : float 

Previous and new values of time. They will be used as a bracketing 

interval. 

 

Returns 

------- 

root : float 

Found solution. 

""" 

from scipy.optimize import brentq 

return brentq(lambda t: event(t, sol(t)), t_old, t, 

xtol=4 * EPS, rtol=4 * EPS) 

 

 

def handle_events(sol, events, active_events, is_terminal, t_old, t): 

"""Helper function to handle events. 

 

Parameters 

---------- 

sol : DenseOutput 

Function ``sol(t)`` which evaluates an ODE solution between `t_old` 

and `t`. 

events : list of callables, length n_events 

Event functions with signatures ``event(t, y)``. 

active_events : ndarray 

Indices of events which occurred. 

is_terminal : ndarray, shape (n_events,) 

Which events are terminal. 

t_old, t : float 

Previous and new values of time. 

 

Returns 

------- 

root_indices : ndarray 

Indices of events which take zero between `t_old` and `t` and before 

a possible termination. 

roots : ndarray 

Values of t at which events occurred. 

terminate : bool 

Whether a terminal event occurred. 

""" 

roots = [] 

for event_index in active_events: 

roots.append(solve_event_equation(events[event_index], sol, t_old, t)) 

 

roots = np.asarray(roots) 

 

if np.any(is_terminal[active_events]): 

if t > t_old: 

order = np.argsort(roots) 

else: 

order = np.argsort(-roots) 

active_events = active_events[order] 

roots = roots[order] 

t = np.nonzero(is_terminal[active_events])[0][0] 

active_events = active_events[:t + 1] 

roots = roots[:t + 1] 

terminate = True 

else: 

terminate = False 

 

return active_events, roots, terminate 

 

 

def find_active_events(g, g_new, direction): 

"""Find which event occurred during an integration step. 

 

Parameters 

---------- 

g, g_new : array_like, shape (n_events,) 

Values of event functions at a current and next points. 

direction : ndarray, shape (n_events,) 

Event "direction" according to the definition in `solve_ivp`. 

 

Returns 

------- 

active_events : ndarray 

Indices of events which occurred during the step. 

""" 

g, g_new = np.asarray(g), np.asarray(g_new) 

up = (g <= 0) & (g_new >= 0) 

down = (g >= 0) & (g_new <= 0) 

either = up | down 

mask = (up & (direction > 0) | 

down & (direction < 0) | 

either & (direction == 0)) 

 

return np.nonzero(mask)[0] 

 

 

def solve_ivp(fun, t_span, y0, method='RK45', t_eval=None, dense_output=False, 

events=None, vectorized=False, **options): 

"""Solve an initial value problem for a system of ODEs. 

 

This function numerically integrates a system of ordinary differential 

equations given an initial value:: 

 

dy / dt = f(t, y) 

y(t0) = y0 

 

Here t is a one-dimensional independent variable (time), y(t) is an 

n-dimensional vector-valued function (state), and an n-dimensional 

vector-valued function f(t, y) determines the differential equations. 

The goal is to find y(t) approximately satisfying the differential 

equations, given an initial value y(t0)=y0. 

 

Some of the solvers support integration in the complex domain, but note that 

for stiff ODE solvers, the right-hand side must be complex-differentiable 

(satisfy Cauchy-Riemann equations [11]_). To solve a problem in the complex 

domain, pass y0 with a complex data type. Another option is always to 

rewrite your problem for real and imaginary parts separately. 

 

Parameters 

---------- 

fun : callable 

Right-hand side of the system. The calling signature is ``fun(t, y)``. 

Here ``t`` is a scalar, and there are two options for the ndarray ``y``: 

It can either have shape (n,); then ``fun`` must return array_like with 

shape (n,). Alternatively it can have shape (n, k); then ``fun`` 

must return an array_like with shape (n, k), i.e. each column 

corresponds to a single column in ``y``. The choice between the two 

options is determined by `vectorized` argument (see below). The 

vectorized implementation allows a faster approximation of the Jacobian 

by finite differences (required for stiff solvers). 

t_span : 2-tuple of floats 

Interval of integration (t0, tf). The solver starts with t=t0 and 

integrates until it reaches t=tf. 

y0 : array_like, shape (n,) 

Initial state. For problems in the complex domain, pass `y0` with a 

complex data type (even if the initial guess is purely real). 

method : string or `OdeSolver`, optional 

Integration method to use: 

 

* 'RK45' (default): Explicit Runge-Kutta method of order 5(4) [1]_. 

The error is controlled assuming accuracy of the fourth-order 

method, but steps are taken using the fifth-order accurate formula 

(local extrapolation is done). A quartic interpolation polynomial 

is used for the dense output [2]_. Can be applied in the complex domain. 

* 'RK23': Explicit Runge-Kutta method of order 3(2) [3]_. The error 

is controlled assuming accuracy of the second-order method, but 

steps are taken using the third-order accurate formula (local 

extrapolation is done). A cubic Hermite polynomial is used for the 

dense output. Can be applied in the complex domain. 

* 'Radau': Implicit Runge-Kutta method of the Radau IIA family of 

order 5 [4]_. The error is controlled with a third-order accurate 

embedded formula. A cubic polynomial which satisfies the 

collocation conditions is used for the dense output. 

* 'BDF': Implicit multi-step variable-order (1 to 5) method based 

on a backward differentiation formula for the derivative 

approximation [5]_. The implementation follows the one described 

in [6]_. A quasi-constant step scheme is used and accuracy is 

enhanced using the NDF modification. Can be applied in the complex 

domain. 

* 'LSODA': Adams/BDF method with automatic stiffness detection and 

switching [7]_, [8]_. This is a wrapper of the Fortran solver 

from ODEPACK. 

 

You should use the 'RK45' or 'RK23' method for non-stiff problems and 

'Radau' or 'BDF' for stiff problems [9]_. If not sure, first try to run 

'RK45'. If needs unusually many iterations, diverges, or fails, your 

problem is likely to be stiff and you should use 'Radau' or 'BDF'. 

'LSODA' can also be a good universal choice, but it might be somewhat 

less convenient to work with as it wraps old Fortran code. 

 

You can also pass an arbitrary class derived from `OdeSolver` which 

implements the solver. 

dense_output : bool, optional 

Whether to compute a continuous solution. Default is False. 

t_eval : array_like or None, optional 

Times at which to store the computed solution, must be sorted and lie 

within `t_span`. If None (default), use points selected by the solver. 

events : callable, list of callables or None, optional 

Types of events to track. Each is defined by a continuous function of 

time and state that becomes zero value in case of an event. Each function 

must have the signature ``event(t, y)`` and return a float. The solver will 

find an accurate value of ``t`` at which ``event(t, y(t)) = 0`` using a 

root-finding algorithm. Additionally each ``event`` function might have 

the following attributes: 

 

* terminal: bool, whether to terminate integration if this 

event occurs. Implicitly False if not assigned. 

* direction: float, direction of a zero crossing. If `direction` 

is positive, `event` must go from negative to positive, and 

vice versa if `direction` is negative. If 0, then either direction 

will count. Implicitly 0 if not assigned. 

 

You can assign attributes like ``event.terminal = True`` to any 

function in Python. If None (default), events won't be tracked. 

vectorized : bool, optional 

Whether `fun` is implemented in a vectorized fashion. Default is False. 

options 

Options passed to a chosen solver. All options available for already 

implemented solvers are listed below. 

max_step : float, optional 

Maximum allowed step size. Default is np.inf, i.e. the step size is not 

bounded and determined solely by the solver. 

rtol, atol : float and array_like, optional 

Relative and absolute tolerances. The solver keeps the local error 

estimates less than ``atol + rtol * abs(y)``. Here `rtol` controls a 

relative accuracy (number of correct digits). But if a component of `y` 

is approximately below `atol`, the error only needs to fall within 

the same `atol` threshold, and the number of correct digits is not 

guaranteed. If components of y have different scales, it might be 

beneficial to set different `atol` values for different components by 

passing array_like with shape (n,) for `atol`. Default values are 

1e-3 for `rtol` and 1e-6 for `atol`. 

jac : {None, array_like, sparse_matrix, callable}, optional 

Jacobian matrix of the right-hand side of the system with respect to 

y, required by the 'Radau', 'BDF' and 'LSODA' method. The Jacobian matrix 

has shape (n, n) and its element (i, j) is equal to ``d f_i / d y_j``. 

There are three ways to define the Jacobian: 

 

* If array_like or sparse_matrix, the Jacobian is assumed to 

be constant. Not supported by 'LSODA'. 

* If callable, the Jacobian is assumed to depend on both 

t and y; it will be called as ``jac(t, y)`` as necessary. 

For the 'Radau' and 'BDF' methods, the return value might be a 

sparse matrix. 

* If None (default), the Jacobian will be approximated by 

finite differences. 

 

It is generally recommended to provide the Jacobian rather than 

relying on a finite-difference approximation. 

jac_sparsity : {None, array_like, sparse matrix}, optional 

Defines a sparsity structure of the Jacobian matrix for a 

finite-difference approximation. Its shape must be (n, n). This argument 

is ignored if `jac` is not `None`. If the Jacobian has only few non-zero 

elements in *each* row, providing the sparsity structure will greatly 

speed up the computations [10]_. A zero entry means that a corresponding 

element in the Jacobian is always zero. If None (default), the Jacobian 

is assumed to be dense. 

Not supported by 'LSODA', see `lband` and `uband` instead. 

lband, uband : int or None 

Parameters defining the bandwidth of the Jacobian for the 'LSODA' method, 

i.e., ``jac[i, j] != 0 only for i - lband <= j <= i + uband``. Setting 

these requires your jac routine to return the Jacobian in the packed format: 

the returned array must have ``n`` columns and ``uband + lband + 1`` 

rows in which Jacobian diagonals are written. Specifically 

``jac_packed[uband + i - j , j] = jac[i, j]``. The same format is used 

in `scipy.linalg.solve_banded` (check for an illustration). 

These parameters can be also used with ``jac=None`` to reduce the 

number of Jacobian elements estimated by finite differences. 

min_step, first_step : float, optional 

The minimum allowed step size and the initial step size respectively 

for 'LSODA' method. By default `min_step` is zero and `first_step` is 

selected automatically. 

 

Returns 

------- 

Bunch object with the following fields defined: 

t : ndarray, shape (n_points,) 

Time points. 

y : ndarray, shape (n, n_points) 

Values of the solution at `t`. 

sol : `OdeSolution` or None 

Found solution as `OdeSolution` instance; None if `dense_output` was 

set to False. 

t_events : list of ndarray or None 

Contains for each event type a list of arrays at which an event of 

that type event was detected. None if `events` was None. 

nfev : int 

Number of evaluations of the right-hand side. 

njev : int 

Number of evaluations of the Jacobian. 

nlu : int 

Number of LU decompositions. 

status : int 

Reason for algorithm termination: 

 

* -1: Integration step failed. 

* 0: The solver successfully reached the end of `tspan`. 

* 1: A termination event occurred. 

 

message : string 

Human-readable description of the termination reason. 

success : bool 

True if the solver reached the interval end or a termination event 

occurred (``status >= 0``). 

 

References 

---------- 

.. [1] J. R. Dormand, P. J. Prince, "A family of embedded Runge-Kutta 

formulae", Journal of Computational and Applied Mathematics, Vol. 6, 

No. 1, pp. 19-26, 1980. 

.. [2] L. W. Shampine, "Some Practical Runge-Kutta Formulas", Mathematics 

of Computation,, Vol. 46, No. 173, pp. 135-150, 1986. 

.. [3] P. Bogacki, L.F. Shampine, "A 3(2) Pair of Runge-Kutta Formulas", 

Appl. Math. Lett. Vol. 2, No. 4. pp. 321-325, 1989. 

.. [4] E. Hairer, G. Wanner, "Solving Ordinary Differential Equations II: 

Stiff and Differential-Algebraic Problems", Sec. IV.8. 

.. [5] `Backward Differentiation Formula 

<https://en.wikipedia.org/wiki/Backward_differentiation_formula>`_ 

on Wikipedia. 

.. [6] L. F. Shampine, M. W. Reichelt, "THE MATLAB ODE SUITE", SIAM J. SCI. 

COMPUTE., Vol. 18, No. 1, pp. 1-22, January 1997. 

.. [7] A. C. Hindmarsh, "ODEPACK, A Systematized Collection of ODE 

Solvers," IMACS Transactions on Scientific Computation, Vol 1., 

pp. 55-64, 1983. 

.. [8] L. Petzold, "Automatic selection of methods for solving stiff and 

nonstiff systems of ordinary differential equations", SIAM Journal 

on Scientific and Statistical Computing, Vol. 4, No. 1, pp. 136-148, 

1983. 

.. [9] `Stiff equation <https://en.wikipedia.org/wiki/Stiff_equation>`_ on 

Wikipedia. 

.. [10] A. Curtis, M. J. D. Powell, and J. Reid, "On the estimation of 

sparse Jacobian matrices", Journal of the Institute of Mathematics 

and its Applications, 13, pp. 117-120, 1974. 

.. [11] `Cauchy-Riemann equations 

<https://en.wikipedia.org/wiki/Cauchy-Riemann_equations>`_ on 

Wikipedia. 

 

Examples 

-------- 

Basic exponential decay showing automatically chosen time points. 

 

>>> from scipy.integrate import solve_ivp 

>>> def exponential_decay(t, y): return -0.5 * y 

>>> sol = solve_ivp(exponential_decay, [0, 10], [2, 4, 8]) 

>>> print(sol.t) 

[ 0. 0.11487653 1.26364188 3.06061781 4.85759374 

6.65456967 8.4515456 10. ] 

>>> print(sol.y) 

[[2. 1.88836035 1.06327177 0.43319312 0.17648948 0.0719045 

0.02929499 0.01350938] 

[4. 3.7767207 2.12654355 0.86638624 0.35297895 0.143809 

0.05858998 0.02701876] 

[8. 7.5534414 4.25308709 1.73277247 0.7059579 0.287618 

0.11717996 0.05403753]] 

 

Specifying points where the solution is desired. 

 

>>> sol = solve_ivp(exponential_decay, [0, 10], [2, 4, 8], 

... t_eval=[0, 1, 2, 4, 10]) 

>>> print(sol.t) 

[ 0 1 2 4 10] 

>>> print(sol.y) 

[[2. 1.21305369 0.73534021 0.27066736 0.01350938] 

[4. 2.42610739 1.47068043 0.54133472 0.02701876] 

[8. 4.85221478 2.94136085 1.08266944 0.05403753]] 

 

Cannon fired upward with terminal event upon impact. The ``terminal`` and 

``direction`` fields of an event are applied by monkey patching a function. 

Here ``y[0]`` is position and ``y[1]`` is velocity. The projectile starts at 

position 0 with velocity +10. Note that the integration never reaches t=100 

because the event is terminal. 

 

>>> def upward_cannon(t, y): return [y[1], -0.5] 

>>> def hit_ground(t, y): return y[1] 

>>> hit_ground.terminal = True 

>>> hit_ground.direction = -1 

>>> sol = solve_ivp(upward_cannon, [0, 100], [0, 10], events=hit_ground) 

>>> print(sol.t_events) 

[array([ 20.])] 

>>> print(sol.t) 

[0.00000000e+00 9.99900010e-05 1.09989001e-03 1.10988901e-02 

1.11088891e-01 1.11098890e+00 1.11099890e+01 2.00000000e+01] 

""" 

if method not in METHODS and not ( 

inspect.isclass(method) and issubclass(method, OdeSolver)): 

raise ValueError("`method` must be one of {} or OdeSolver class." 

.format(METHODS)) 

 

t0, tf = float(t_span[0]), float(t_span[1]) 

 

if t_eval is not None: 

t_eval = np.asarray(t_eval) 

if t_eval.ndim != 1: 

raise ValueError("`t_eval` must be 1-dimensional.") 

 

if np.any(t_eval < min(t0, tf)) or np.any(t_eval > max(t0, tf)): 

raise ValueError("Values in `t_eval` are not within `t_span`.") 

 

d = np.diff(t_eval) 

if tf > t0 and np.any(d <= 0) or tf < t0 and np.any(d >= 0): 

raise ValueError("Values in `t_eval` are not properly sorted.") 

 

if tf > t0: 

t_eval_i = 0 

else: 

# Make order of t_eval decreasing to use np.searchsorted. 

t_eval = t_eval[::-1] 

# This will be an upper bound for slices. 

t_eval_i = t_eval.shape[0] 

 

if method in METHODS: 

method = METHODS[method] 

 

solver = method(fun, t0, y0, tf, vectorized=vectorized, **options) 

 

if t_eval is None: 

ts = [t0] 

ys = [y0] 

else: 

ts = [] 

ys = [] 

 

interpolants = [] 

 

events, is_terminal, event_dir = prepare_events(events) 

 

if events is not None: 

g = [event(t0, y0) for event in events] 

t_events = [[] for _ in range(len(events))] 

else: 

t_events = None 

 

status = None 

while status is None: 

message = solver.step() 

 

if solver.status == 'finished': 

status = 0 

elif solver.status == 'failed': 

status = -1 

break 

 

t_old = solver.t_old 

t = solver.t 

y = solver.y 

 

if dense_output: 

sol = solver.dense_output() 

interpolants.append(sol) 

else: 

sol = None 

 

if events is not None: 

g_new = [event(t, y) for event in events] 

active_events = find_active_events(g, g_new, event_dir) 

if active_events.size > 0: 

if sol is None: 

sol = solver.dense_output() 

 

root_indices, roots, terminate = handle_events( 

sol, events, active_events, is_terminal, t_old, t) 

 

for e, te in zip(root_indices, roots): 

t_events[e].append(te) 

 

if terminate: 

status = 1 

t = roots[-1] 

y = sol(t) 

 

g = g_new 

 

if t_eval is None: 

ts.append(t) 

ys.append(y) 

else: 

# The value in t_eval equal to t will be included. 

if solver.direction > 0: 

t_eval_i_new = np.searchsorted(t_eval, t, side='right') 

t_eval_step = t_eval[t_eval_i:t_eval_i_new] 

else: 

t_eval_i_new = np.searchsorted(t_eval, t, side='left') 

# It has to be done with two slice operations, because 

# you can't slice to 0-th element inclusive using backward 

# slicing. 

t_eval_step = t_eval[t_eval_i_new:t_eval_i][::-1] 

 

if t_eval_step.size > 0: 

if sol is None: 

sol = solver.dense_output() 

ts.append(t_eval_step) 

ys.append(sol(t_eval_step)) 

t_eval_i = t_eval_i_new 

 

message = MESSAGES.get(status, message) 

 

if t_events is not None: 

t_events = [np.asarray(te) for te in t_events] 

 

if t_eval is None: 

ts = np.array(ts) 

ys = np.vstack(ys).T 

else: 

ts = np.hstack(ts) 

ys = np.hstack(ys) 

 

if dense_output: 

sol = OdeSolution(ts, interpolants) 

else: 

sol = None 

 

return OdeResult(t=ts, y=ys, sol=sol, t_events=t_events, nfev=solver.nfev, 

njev=solver.njev, nlu=solver.nlu, status=status, 

message=message, success=status >= 0)