Undocumented
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
No class docstring; 0/1 class variable, 1/7 method documented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
No class docstring; 2/3 methods documented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Class |
|
Undocumented |
Function | assert |
Undocumented |
Function | assert |
Undocumented |
Function | assert |
Undocumented |
Function | assert |
Undocumented |
Function | assert |
Undocumented |
Function | assert |
Undocumented |
Function | assert |
Undocumented |
Function | bad |
The blocklisted trig functions are not accurate on aarch64/PPC for complex256. Rather than dig through the actual problem skip the test. This should be fixed when we can move past glibc2.17 which is the version in manylinux2014... |
Function | floor |
Undocumented |
Function | interesting |
Helper to create "interesting" operands to cover common code paths: * scalar inputs * only first "values" is an array (e.g. scalar division fast-paths) * Longer array (SIMD) placing the value of interest at different positions * Oddly strided arrays which may not be SIMD compatible... |
Function | on |
True if we are running on a Power PC platform. |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Check np.nan is a positive nan. |
Function | test |
Test bug in reduceat when structured arrays are not copied. |
Function | test |
Reduceat should work with empty arrays |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Function | test |
Undocumented |
Constant | UFUNCS |
Undocumented |
Constant | UFUNCS |
Undocumented |
Constant | UFUNCS |
Undocumented |
Constant | UFUNCS |
Undocumented |
Constant | UFUNCS |
Undocumented |
Variable | avx |
Undocumented |
Class | _ |
Undocumented |
Function | _check |
Check for a branch cut in a function. |
Function | _signs |
Undocumented |
Function | _test |
Undocumented |
Function | _test |
Undocumented |
The blocklisted trig functions are not accurate on aarch64/PPC for complex256. Rather than dig through the actual problem skip the test. This should be fixed when we can move past glibc2.17 which is the version in manylinux2014
Helper to create "interesting" operands to cover common code paths: * scalar inputs * only first "values" is an array (e.g. scalar division fast-paths) * Longer array (SIMD) placing the value of interest at different positions * Oddly strided arrays which may not be SIMD compatible
It does not attempt to cover unaligned access or mixed dtypes. These are normally handled by the casting/buffering machinery.
This is not a fixture (currently), since I believe a fixture normally only yields once?
def test_bad_legacy_gufunc_silent_errors(x1): (source) ¶
Undocumented
def test_memoverlap_accumulate(ftype): (source) ¶
Undocumented
def test_memoverlap_accumulate_cmp(ufunc, dtype): (source) ¶
Undocumented
def test_memoverlap_accumulate_symmetric(ufunc, dtype): (source) ¶
Undocumented
@pytest.mark.xfail(
def test_nextafterl(): (source) ¶
Undocumented
def test_outer_subclass_preserve(arr): (source) ¶
Undocumented
@pytest.mark.xfail(
def test_spacingl(): (source) ¶
Undocumented
Check for a branch cut in a function.
Assert that x0
lies on a branch cut of function f
and f
is
continuous from the direction dx
.
Parameters | |
f:func | Function to check |
x0:array-like | Point on branch cut |
dx:array-like | Direction to check continuity in |
re | Change of sign of the real or imaginary part expected |
im | Change of sign of the real or imaginary part expected |
sigbool | Whether to check if the branch cut respects signed zero (if applicable) |
dtype:dtype | Dtype to check (should be complex) |