Skip to content

🐛 integrate: several BDF fixes #882

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Sep 13, 2025
+56 −28
Merged

🐛 integrate: several BDF fixes #882

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion scipy-stubs/integrate/_ivp/base.pyi
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,7 @@ import numpy as np
import optype.numpy as onp
import optype.numpy.compat as npc

_ScalarT = TypeVar("_ScalarT", bound=np.float64 | np.complex128, default=np.float64)
_ScalarT = TypeVar("_ScalarT", bound=np.float64 | np.complex128, default=np.float64 | Any)
_ScalarT_co = TypeVar("_ScalarT_co", bound=npc.inexact, default=np.float64 | Any, covariant=True)

_ToFunReal: TypeAlias = Callable[[float, onp.ArrayND[np.float64]], onp.ToFloatND]
Expand Down
82 changes: 55 additions & 27 deletions scipy-stubs/integrate/_ivp/bdf.pyi
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,23 +1,28 @@
from collections.abc import Callable
from typing import Final, Generic, Never, TypeAlias
from typing import Any, Final, Generic, Never, TypeAlias, overload
from typing_extensions import TypeVar

import numpy as np
import optype.numpy as onp
import optype.numpy.compat as npc

from .base import DenseOutput, OdeSolver
from scipy.sparse import sparray, spmatrix
from scipy.sparse import csc_matrix, sparray, spmatrix

###

_SCT_co = TypeVar("_SCT_co", covariant=True, bound=npc.inexact, default=np.float64 | np.complex128)
_NumberT = TypeVar("_NumberT", bound=npc.number)
_InexactT = TypeVar("_InexactT", bound=np.float64 | np.complex128, default=np.float64 | Any)

_LU: TypeAlias = tuple[onp.ArrayND[npc.inexact], onp.ArrayND[npc.integer]]
_FuncLU: TypeAlias = Callable[[onp.ArrayND[np.float64]], _LU] | Callable[[onp.ArrayND[np.complex128]], _LU]
_FuncSolveLU: TypeAlias = Callable[[_LU, onp.ArrayND], onp.ArrayND[npc.inexact]]

_ToJac: TypeAlias = onp.ToComplex2D | spmatrix | sparray
_Sparse2D: TypeAlias = spmatrix[_NumberT] | sparray[_NumberT, tuple[int, int]]
_ArrayOrCSC: TypeAlias = onp.Array2D[_NumberT] | csc_matrix[_NumberT]

_ToJacReal: TypeAlias = onp.ToFloat2D | _Sparse2D[npc.floating | npc.integer]
_ToJacComplex: TypeAlias = onp.ToComplex2D | _Sparse2D[npc.number]

###

Expand All @@ -26,40 +31,63 @@ NEWTON_MAXITER: Final = 4
MIN_FACTOR: Final = 0.2
MAX_FACTOR: Final = 10

class BDF(OdeSolver, Generic[_SCT_co]):
class BDF(OdeSolver[_InexactT], Generic[_InexactT]):
max_step: float
h_abs: float
h_abs_old: float | None
error_norm_old: None
newton_tol: float
jac_factor: onp.ArrayND[np.float64] | None # 1d

LU: _LU
jac_factor: onp.Array1D[np.float64] | None
jac: Callable[[float, onp.ArrayND[_InexactT]], _ArrayOrCSC[_InexactT]] | None

J: _ArrayOrCSC[_InexactT]
I: _ArrayOrCSC[_InexactT]
D: onp.Array2D[_InexactT]

LU: _LU | None
lu: _FuncLU
solve_lu: _FuncSolveLU

I: onp.ArrayND[_SCT_co]
error_const: onp.ArrayND[np.float64]
gamma: onp.ArrayND[np.float64]
alpha: onp.ArrayND[np.float64]
D: onp.ArrayND[np.float64]
order: int
gamma: onp.Array1D[np.float64]
alpha: onp.Array1D[np.float64]
error_const: onp.Array1D[np.float64]

order: int | np.intp
n_equal_steps: int

@overload
def __init__(
self: OdeSolver[np.float64],
/,
fun: Callable[[float, onp.ArrayND[np.float64]], onp.ToFloatND],
t0: float,
y0: onp.ToFloatND,
t_bound: float,
max_step: float = ..., # = np.inf
rtol: float = 1e-3,
atol: float = 1e-6,
jac: _ToJacReal | Callable[[float, onp.ArrayND[np.float64]], _ToJacReal] | None = None,
jac_sparsity: _ToJacReal | None = None,
vectorized: bool = False,
first_step: float | None = None,
**extraneous: Never,
) -> None: ...
@overload
def __init__(
self,
self: OdeSolver[np.complex128],
/,
fun: Callable[[float, onp.Array1D[_SCT_co]], onp.ToComplex1D],
t0: onp.ToFloat,
y0: onp.Array1D[_SCT_co] | onp.ToComplexND,
t_bound: onp.ToFloat,
max_step: onp.ToFloat = ...,
rtol: onp.ToFloat = 0.001,
atol: onp.ToFloat = 1e-06,
jac: _ToJac | Callable[[float, onp.ArrayND[_SCT_co]], _ToJac] | None = None,
jac_sparsity: _ToJac | None = None,
fun: Callable[[float, onp.ArrayND[np.complex128]], onp.ToComplexND],
t0: float,
y0: onp.ToJustComplexND,
t_bound: float,
max_step: float = ..., # = np.inf
rtol: float = 1e-3,
atol: float = 1e-6,
jac: _ToJacComplex | Callable[[float, onp.ArrayND[np.complex128]], _ToJacComplex] | None = None,
jac_sparsity: _ToJacComplex | None = None,
vectorized: bool = False,
first_step: onp.ToFloat | None = None,
first_step: float | None = None,
**extraneous: Never,
) -> None: ...

Expand All @@ -73,13 +101,13 @@ class BdfDenseOutput(DenseOutput[np.float64]):
def compute_R(order: int, factor: float) -> onp.ArrayND[np.float64]: ...
def change_D(D: onp.ArrayND[np.float64], order: int, factor: float) -> None: ...
def solve_bdf_system(
fun: Callable[[float, onp.ArrayND[_SCT_co]], onp.ToComplex1D],
fun: Callable[[float, onp.ArrayND[_InexactT]], onp.ToComplex1D],
t_new: onp.ToFloat,
y_predict: onp.ArrayND[_SCT_co],
y_predict: onp.ArrayND[_InexactT],
c: float,
psi: onp.ArrayND[np.float64],
LU: _FuncLU,
solve_lu: _FuncSolveLU,
scale: onp.ArrayND[np.float64],
tol: float,
) -> tuple[bool, int, onp.ArrayND[_SCT_co], onp.ArrayND[_SCT_co]]: ...
) -> tuple[bool, int, onp.ArrayND[_InexactT], onp.ArrayND[_InexactT]]: ...