gitignore entfernt.

2024-07-12 10:46:49 +02:00
parent 6e015d30a1
commit e0d1fc976a
4887 changed files with 1191510 additions and 0 deletions
@@ -0,0 +1,80 @@
+"""
+``numpy.linalg``
+================
+
+The NumPy linear algebra functions rely on BLAS and LAPACK to provide efficient
+low level implementations of standard linear algebra algorithms. Those
+libraries may be provided by NumPy itself using C versions of a subset of their
+reference implementations but, when possible, highly optimized libraries that
+take advantage of specialized processor functionality are preferred. Examples
+of such libraries are OpenBLAS, MKL (TM), and ATLAS. Because those libraries
+are multithreaded and processor dependent, environmental variables and external
+packages such as threadpoolctl may be needed to control the number of threads
+or specify the processor architecture.
+
+- OpenBLAS: https://www.openblas.net/
+- threadpoolctl: https://github.com/joblib/threadpoolctl
+
+Please note that the most-used linear algebra functions in NumPy are present in
+the main ``numpy`` namespace rather than in ``numpy.linalg``.  There are:
+``dot``, ``vdot``, ``inner``, ``outer``, ``matmul``, ``tensordot``, ``einsum``,
+``einsum_path`` and ``kron``.
+
+Functions present in numpy.linalg are listed below.
+
+
+Matrix and vector products
+--------------------------
+
+   multi_dot
+   matrix_power
+
+Decompositions
+--------------
+
+   cholesky
+   qr
+   svd
+
+Matrix eigenvalues
+------------------
+
+   eig
+   eigh
+   eigvals
+   eigvalsh
+
+Norms and other numbers
+-----------------------
+
+   norm
+   cond
+   det
+   matrix_rank
+   slogdet
+
+Solving equations and inverting matrices
+----------------------------------------
+
+   solve
+   tensorsolve
+   lstsq
+   inv
+   pinv
+   tensorinv
+
+Exceptions
+----------
+
+   LinAlgError
+
+"""
+# To get sub-modules
+from . import linalg
+from .linalg import *
+
+__all__ = linalg.__all__.copy()
+
+from numpy._pytesttester import PytestTester
+test = PytestTester(__name__)
+del PytestTester
@@ -0,0 +1,30 @@
+from numpy.linalg.linalg import (
+    matrix_power as matrix_power,
+    solve as solve,
+    tensorsolve as tensorsolve,
+    tensorinv as tensorinv,
+    inv as inv,
+    cholesky as cholesky,
+    eigvals as eigvals,
+    eigvalsh as eigvalsh,
+    pinv as pinv,
+    slogdet as slogdet,
+    det as det,
+    svd as svd,
+    eig as eig,
+    eigh as eigh,
+    lstsq as lstsq,
+    norm as norm,
+    qr as qr,
+    cond as cond,
+    matrix_rank as matrix_rank,
+    multi_dot as multi_dot,
+)
+
+from numpy._pytesttester import PytestTester
+
+__all__: list[str]
+__path__: list[str]
+test: PytestTester
+
+class LinAlgError(Exception): ...
@@ -0,0 +1,297 @@
+from collections.abc import Iterable
+from typing import (
+    Literal as L,
+    overload,
+    TypeVar,
+    Any,
+    SupportsIndex,
+    SupportsInt,
+    NamedTuple,
+    Generic,
+)
+
+from numpy import (
+    generic,
+    floating,
+    complexfloating,
+    int32,
+    float64,
+    complex128,
+)
+
+from numpy.linalg import LinAlgError as LinAlgError
+
+from numpy._typing import (
+    NDArray,
+    ArrayLike,
+    _ArrayLikeInt_co,
+    _ArrayLikeFloat_co,
+    _ArrayLikeComplex_co,
+    _ArrayLikeTD64_co,
+    _ArrayLikeObject_co,
+)
+
+_T = TypeVar("_T")
+_ArrayType = TypeVar("_ArrayType", bound=NDArray[Any])
+_SCT = TypeVar("_SCT", bound=generic, covariant=True)
+_SCT2 = TypeVar("_SCT2", bound=generic, covariant=True)
+
+_2Tuple = tuple[_T, _T]
+_ModeKind = L["reduced", "complete", "r", "raw"]
+
+__all__: list[str]
+
+class EigResult(NamedTuple):
+    eigenvalues: NDArray[Any]
+    eigenvectors: NDArray[Any]
+
+class EighResult(NamedTuple):
+    eigenvalues: NDArray[Any]
+    eigenvectors: NDArray[Any]
+
+class QRResult(NamedTuple):
+    Q: NDArray[Any]
+    R: NDArray[Any]
+
+class SlogdetResult(NamedTuple):
+    # TODO: `sign` and `logabsdet` are scalars for input 2D arrays and
+    # a `(x.ndim - 2)`` dimensionl arrays otherwise
+    sign: Any
+    logabsdet: Any
+
+class SVDResult(NamedTuple):
+    U: NDArray[Any]
+    S: NDArray[Any]
+    Vh: NDArray[Any]
+
+@overload
+def tensorsolve(
+    a: _ArrayLikeInt_co,
+    b: _ArrayLikeInt_co,
+    axes: None | Iterable[int] =...,
+) -> NDArray[float64]: ...
+@overload
+def tensorsolve(
+    a: _ArrayLikeFloat_co,
+    b: _ArrayLikeFloat_co,
+    axes: None | Iterable[int] =...,
+) -> NDArray[floating[Any]]: ...
+@overload
+def tensorsolve(
+    a: _ArrayLikeComplex_co,
+    b: _ArrayLikeComplex_co,
+    axes: None | Iterable[int] =...,
+) -> NDArray[complexfloating[Any, Any]]: ...
+
+@overload
+def solve(
+    a: _ArrayLikeInt_co,
+    b: _ArrayLikeInt_co,
+) -> NDArray[float64]: ...
+@overload
+def solve(
+    a: _ArrayLikeFloat_co,
+    b: _ArrayLikeFloat_co,
+) -> NDArray[floating[Any]]: ...
+@overload
+def solve(
+    a: _ArrayLikeComplex_co,
+    b: _ArrayLikeComplex_co,
+) -> NDArray[complexfloating[Any, Any]]: ...
+
+@overload
+def tensorinv(
+    a: _ArrayLikeInt_co,
+    ind: int = ...,
+) -> NDArray[float64]: ...
+@overload
+def tensorinv(
+    a: _ArrayLikeFloat_co,
+    ind: int = ...,
+) -> NDArray[floating[Any]]: ...
+@overload
+def tensorinv(
+    a: _ArrayLikeComplex_co,
+    ind: int = ...,
+) -> NDArray[complexfloating[Any, Any]]: ...
+
+@overload
+def inv(a: _ArrayLikeInt_co) -> NDArray[float64]: ...
+@overload
+def inv(a: _ArrayLikeFloat_co) -> NDArray[floating[Any]]: ...
+@overload
+def inv(a: _ArrayLikeComplex_co) -> NDArray[complexfloating[Any, Any]]: ...
+
+# TODO: The supported input and output dtypes are dependent on the value of `n`.
+# For example: `n < 0` always casts integer types to float64
+def matrix_power(
+    a: _ArrayLikeComplex_co | _ArrayLikeObject_co,
+    n: SupportsIndex,
+) -> NDArray[Any]: ...
+
+@overload
+def cholesky(a: _ArrayLikeInt_co) -> NDArray[float64]: ...
+@overload
+def cholesky(a: _ArrayLikeFloat_co) -> NDArray[floating[Any]]: ...
+@overload
+def cholesky(a: _ArrayLikeComplex_co) -> NDArray[complexfloating[Any, Any]]: ...
+
+@overload
+def qr(a: _ArrayLikeInt_co, mode: _ModeKind = ...) -> QRResult: ...
+@overload
+def qr(a: _ArrayLikeFloat_co, mode: _ModeKind = ...) -> QRResult: ...
+@overload
+def qr(a: _ArrayLikeComplex_co, mode: _ModeKind = ...) -> QRResult: ...
+
+@overload
+def eigvals(a: _ArrayLikeInt_co) -> NDArray[float64] | NDArray[complex128]: ...
+@overload
+def eigvals(a: _ArrayLikeFloat_co) -> NDArray[floating[Any]] | NDArray[complexfloating[Any, Any]]: ...
+@overload
+def eigvals(a: _ArrayLikeComplex_co) -> NDArray[complexfloating[Any, Any]]: ...
+
+@overload
+def eigvalsh(a: _ArrayLikeInt_co, UPLO: L["L", "U", "l", "u"] = ...) -> NDArray[float64]: ...
+@overload
+def eigvalsh(a: _ArrayLikeComplex_co, UPLO: L["L", "U", "l", "u"] = ...) -> NDArray[floating[Any]]: ...
+
+@overload
+def eig(a: _ArrayLikeInt_co) -> EigResult: ...
+@overload
+def eig(a: _ArrayLikeFloat_co) -> EigResult: ...
+@overload
+def eig(a: _ArrayLikeComplex_co) -> EigResult: ...
+
+@overload
+def eigh(
+    a: _ArrayLikeInt_co,
+    UPLO: L["L", "U", "l", "u"] = ...,
+) -> EighResult: ...
+@overload
+def eigh(
+    a: _ArrayLikeFloat_co,
+    UPLO: L["L", "U", "l", "u"] = ...,
+) -> EighResult: ...
+@overload
+def eigh(
+    a: _ArrayLikeComplex_co,
+    UPLO: L["L", "U", "l", "u"] = ...,
+) -> EighResult: ...
+
+@overload
+def svd(
+    a: _ArrayLikeInt_co,
+    full_matrices: bool = ...,
+    compute_uv: L[True] = ...,
+    hermitian: bool = ...,
+) -> SVDResult: ...
+@overload
+def svd(
+    a: _ArrayLikeFloat_co,
+    full_matrices: bool = ...,
+    compute_uv: L[True] = ...,
+    hermitian: bool = ...,
+) -> SVDResult: ...
+@overload
+def svd(
+    a: _ArrayLikeComplex_co,
+    full_matrices: bool = ...,
+    compute_uv: L[True] = ...,
+    hermitian: bool = ...,
+) -> SVDResult: ...
+@overload
+def svd(
+    a: _ArrayLikeInt_co,
+    full_matrices: bool = ...,
+    compute_uv: L[False] = ...,
+    hermitian: bool = ...,
+) -> NDArray[float64]: ...
+@overload
+def svd(
+    a: _ArrayLikeComplex_co,
+    full_matrices: bool = ...,
+    compute_uv: L[False] = ...,
+    hermitian: bool = ...,
+) -> NDArray[floating[Any]]: ...
+
+# TODO: Returns a scalar for 2D arrays and
+# a `(x.ndim - 2)`` dimensionl array otherwise
+def cond(x: _ArrayLikeComplex_co, p: None | float | L["fro", "nuc"] = ...) -> Any: ...
+
+# TODO: Returns `int` for <2D arrays and `intp` otherwise
+def matrix_rank(
+    A: _ArrayLikeComplex_co,
+    tol: None | _ArrayLikeFloat_co = ...,
+    hermitian: bool = ...,
+) -> Any: ...
+
+@overload
+def pinv(
+    a: _ArrayLikeInt_co,
+    rcond: _ArrayLikeFloat_co = ...,
+    hermitian: bool = ...,
+) -> NDArray[float64]: ...
+@overload
+def pinv(
+    a: _ArrayLikeFloat_co,
+    rcond: _ArrayLikeFloat_co = ...,
+    hermitian: bool = ...,
+) -> NDArray[floating[Any]]: ...
+@overload
+def pinv(
+    a: _ArrayLikeComplex_co,
+    rcond: _ArrayLikeFloat_co = ...,
+    hermitian: bool = ...,
+) -> NDArray[complexfloating[Any, Any]]: ...
+
+# TODO: Returns a 2-tuple of scalars for 2D arrays and
+# a 2-tuple of `(a.ndim - 2)`` dimensionl arrays otherwise
+def slogdet(a: _ArrayLikeComplex_co) -> SlogdetResult: ...
+
+# TODO: Returns a 2-tuple of scalars for 2D arrays and
+# a 2-tuple of `(a.ndim - 2)`` dimensionl arrays otherwise
+def det(a: _ArrayLikeComplex_co) -> Any: ...
+
+@overload
+def lstsq(a: _ArrayLikeInt_co, b: _ArrayLikeInt_co, rcond: None | float = ...) -> tuple[
+    NDArray[float64],
+    NDArray[float64],
+    int32,
+    NDArray[float64],
+]: ...
+@overload
+def lstsq(a: _ArrayLikeFloat_co, b: _ArrayLikeFloat_co, rcond: None | float = ...) -> tuple[
+    NDArray[floating[Any]],
+    NDArray[floating[Any]],
+    int32,
+    NDArray[floating[Any]],
+]: ...
+@overload
+def lstsq(a: _ArrayLikeComplex_co, b: _ArrayLikeComplex_co, rcond: None | float = ...) -> tuple[
+    NDArray[complexfloating[Any, Any]],
+    NDArray[floating[Any]],
+    int32,
+    NDArray[floating[Any]],
+]: ...
+
+@overload
+def norm(
+    x: ArrayLike,
+    ord: None | float | L["fro", "nuc"] = ...,
+    axis: None = ...,
+    keepdims: bool = ...,
+) -> floating[Any]: ...
+@overload
+def norm(
+    x: ArrayLike,
+    ord: None | float | L["fro", "nuc"] = ...,
+    axis: SupportsInt | SupportsIndex | tuple[int, ...] = ...,
+    keepdims: bool = ...,
+) -> Any: ...
+
+# TODO: Returns a scalar or array
+def multi_dot(
+    arrays: Iterable[_ArrayLikeComplex_co | _ArrayLikeObject_co | _ArrayLikeTD64_co],
+    *,
+    out: None | NDArray[Any] = ...,
+) -> Any: ...
@@ -0,0 +1,20 @@
+"""Test deprecation and future warnings.
+
+"""
+import numpy as np
+from numpy.testing import assert_warns
+
+
+def test_qr_mode_full_future_warning():
+    """Check mode='full' FutureWarning.
+
+    In numpy 1.8 the mode options 'full' and 'economic' in linalg.qr were
+    deprecated. The release date will probably be sometime in the summer
+    of 2013.
+
+    """
+    a = np.eye(2)
+    assert_warns(DeprecationWarning, np.linalg.qr, a, mode='full')
+    assert_warns(DeprecationWarning, np.linalg.qr, a, mode='f')
+    assert_warns(DeprecationWarning, np.linalg.qr, a, mode='economic')
+    assert_warns(DeprecationWarning, np.linalg.qr, a, mode='e')
@@ -0,0 +1,145 @@
+""" Test functions for linalg module
+"""
+import warnings
+
+import numpy as np
+from numpy import linalg, arange, float64, array, dot, transpose
+from numpy.testing import (
+    assert_, assert_raises, assert_equal, assert_array_equal,
+    assert_array_almost_equal, assert_array_less
+)
+
+
+class TestRegression:
+
+    def test_eig_build(self):
+        # Ticket #652
+        rva = array([1.03221168e+02 + 0.j,
+                     -1.91843603e+01 + 0.j,
+                     -6.04004526e-01 + 15.84422474j,
+                     -6.04004526e-01 - 15.84422474j,
+                     -1.13692929e+01 + 0.j,
+                     -6.57612485e-01 + 10.41755503j,
+                     -6.57612485e-01 - 10.41755503j,
+                     1.82126812e+01 + 0.j,
+                     1.06011014e+01 + 0.j,
+                     7.80732773e+00 + 0.j,
+                     -7.65390898e-01 + 0.j,
+                     1.51971555e-15 + 0.j,
+                     -1.51308713e-15 + 0.j])
+        a = arange(13 * 13, dtype=float64)
+        a.shape = (13, 13)
+        a = a % 17
+        va, ve = linalg.eig(a)
+        va.sort()
+        rva.sort()
+        assert_array_almost_equal(va, rva)
+
+    def test_eigh_build(self):
+        # Ticket 662.
+        rvals = [68.60568999, 89.57756725, 106.67185574]
+
+        cov = array([[77.70273908,   3.51489954,  15.64602427],
+                     [3.51489954,  88.97013878,  -1.07431931],
+                     [15.64602427,  -1.07431931,  98.18223512]])
+
+        vals, vecs = linalg.eigh(cov)
+        assert_array_almost_equal(vals, rvals)
+
+    def test_svd_build(self):
+        # Ticket 627.
+        a = array([[0., 1.], [1., 1.], [2., 1.], [3., 1.]])
+        m, n = a.shape
+        u, s, vh = linalg.svd(a)
+
+        b = dot(transpose(u[:, n:]), a)
+
+        assert_array_almost_equal(b, np.zeros((2, 2)))
+
+    def test_norm_vector_badarg(self):
+        # Regression for #786: Frobenius norm for vectors raises
+        # ValueError.
+        assert_raises(ValueError, linalg.norm, array([1., 2., 3.]), 'fro')
+
+    def test_lapack_endian(self):
+        # For bug #1482
+        a = array([[5.7998084,  -2.1825367],
+                   [-2.1825367,   9.85910595]], dtype='>f8')
+        b = array(a, dtype='<f8')
+
+        ap = linalg.cholesky(a)
+        bp = linalg.cholesky(b)
+        assert_array_equal(ap, bp)
+
+    def test_large_svd_32bit(self):
+        # See gh-4442, 64bit would require very large/slow matrices.
+        x = np.eye(1000, 66)
+        np.linalg.svd(x)
+
+    def test_svd_no_uv(self):
+        # gh-4733
+        for shape in (3, 4), (4, 4), (4, 3):
+            for t in float, complex:
+                a = np.ones(shape, dtype=t)
+                w = linalg.svd(a, compute_uv=False)
+                c = np.count_nonzero(np.absolute(w) > 0.5)
+                assert_equal(c, 1)
+                assert_equal(np.linalg.matrix_rank(a), 1)
+                assert_array_less(1, np.linalg.norm(a, ord=2))
+
+    def test_norm_object_array(self):
+        # gh-7575
+        testvector = np.array([np.array([0, 1]), 0, 0], dtype=object)
+
+        norm = linalg.norm(testvector)
+        assert_array_equal(norm, [0, 1])
+        assert_(norm.dtype == np.dtype('float64'))
+
+        norm = linalg.norm(testvector, ord=1)
+        assert_array_equal(norm, [0, 1])
+        assert_(norm.dtype != np.dtype('float64'))
+
+        norm = linalg.norm(testvector, ord=2)
+        assert_array_equal(norm, [0, 1])
+        assert_(norm.dtype == np.dtype('float64'))
+
+        assert_raises(ValueError, linalg.norm, testvector, ord='fro')
+        assert_raises(ValueError, linalg.norm, testvector, ord='nuc')
+        assert_raises(ValueError, linalg.norm, testvector, ord=np.inf)
+        assert_raises(ValueError, linalg.norm, testvector, ord=-np.inf)
+        assert_raises(ValueError, linalg.norm, testvector, ord=0)
+        assert_raises(ValueError, linalg.norm, testvector, ord=-1)
+        assert_raises(ValueError, linalg.norm, testvector, ord=-2)
+
+        testmatrix = np.array([[np.array([0, 1]), 0, 0],
+                               [0,                0, 0]], dtype=object)
+
+        norm = linalg.norm(testmatrix)
+        assert_array_equal(norm, [0, 1])
+        assert_(norm.dtype == np.dtype('float64'))
+
+        norm = linalg.norm(testmatrix, ord='fro')
+        assert_array_equal(norm, [0, 1])
+        assert_(norm.dtype == np.dtype('float64'))
+
+        assert_raises(TypeError, linalg.norm, testmatrix, ord='nuc')
+        assert_raises(ValueError, linalg.norm, testmatrix, ord=np.inf)
+        assert_raises(ValueError, linalg.norm, testmatrix, ord=-np.inf)
+        assert_raises(ValueError, linalg.norm, testmatrix, ord=0)
+        assert_raises(ValueError, linalg.norm, testmatrix, ord=1)
+        assert_raises(ValueError, linalg.norm, testmatrix, ord=-1)
+        assert_raises(TypeError, linalg.norm, testmatrix, ord=2)
+        assert_raises(TypeError, linalg.norm, testmatrix, ord=-2)
+        assert_raises(ValueError, linalg.norm, testmatrix, ord=3)
+
+    def test_lstsq_complex_larger_rhs(self):
+        # gh-9891
+        size = 20
+        n_rhs = 70
+        G = np.random.randn(size, size) + 1j * np.random.randn(size, size)
+        u = np.random.randn(size, n_rhs) + 1j * np.random.randn(size, n_rhs)
+        b = G.dot(u)
+        # This should work without segmentation fault.
+        u_lstsq, res, rank, sv = linalg.lstsq(G, b, rcond=None)
+        # check results just in case
+        assert_array_almost_equal(u_lstsq, u)